From a38c68c4dd79b331b52438dbd22221e90c646132 Mon Sep 17 00:00:00 2001
From: Ryan Mattson <contact@dalinicus.com>
Date: Fri, 9 Jan 2026 17:28:07 -0600
Subject: [PATCH 1/7] stash

---
 src/processing/audio-processor.ts   | 46 ++++++++++++++++------
 src/processing/index.ts             |  2 +
 src/processing/workflow-chaining.ts | 51 +++++++++++++++++++++---
 src/processing/workflow-executor.ts | 45 +++++----------------
 src/providers/base-provider.ts      | 39 ++++++++++++------
 src/providers/types.ts              | 22 +++++++----
 src/settings/additional-settings.ts | 33 ++++++----------
 src/settings/index.ts               | 32 +++++++++++++--
 src/settings/types.ts               | 12 +++++-
 src/settings/workflows.ts           | 61 +++++++++++++++++++++--------
 src/tokens/types.ts                 | 54 +++++++++++++++++++------
 styles.css                          | 25 +++++++-----
 12 files changed, 283 insertions(+), 139 deletions(-)

diff --git a/src/processing/audio-processor.ts b/src/processing/audio-processor.ts
index 15d7d13..7bc0dd7 100644
--- a/src/processing/audio-processor.ts
+++ b/src/processing/audio-processor.ts
@@ -1,12 +1,32 @@
 import { Buffer } from 'buffer';
 import * as fs from 'fs';
+import { TimestampGranularity } from '../settings';
 
 /**
- * Interface for transcription API response
+ * Word-level timestamp data from transcription API
+ */
+export interface TranscriptionWord {
+    word: string;
+    start: number;
+    end: number;
+}
+
+/**
+ * Segment-level timestamp data from transcription API
+ */
+export interface TranscriptionSegment {
+    text: string;
+    start: number;
+    end: number;
+}
+
+/**
+ * Interface for transcription API response (verbose_json format)
  */
 export interface TranscriptionApiResponse {
     text: string;
-    segments?: Array<{ text: string; start: number; end: number }>;
+    segments?: TranscriptionSegment[];
+    words?: TranscriptionWord[];
 }
 
 /**
@@ -41,7 +61,7 @@ export interface MultipartFormDataOptions {
     boundary: string;
     audioBuffer: Buffer;
     fileName: string;
-    includeTimestamps: boolean;
+    timestampGranularity: TimestampGranularity;
     language?: string;
     additionalFields?: FormField[];
 }
@@ -51,7 +71,7 @@ export interface MultipartFormDataOptions {
  */
 export interface PrepareAudioFormDataOptions {
     audioFilePath: string;
-    includeTimestamps: boolean;
+    timestampGranularity: TimestampGranularity;
     language?: string;
     additionalFields?: FormField[];
 }
@@ -93,12 +113,14 @@ function combinePartsToArrayBuffer(parts: (string | Buffer)[]): ArrayBuffer {
  * Builds multipart form data for audio transcription API requests.
  * Creates a properly formatted multipart/form-data body with the audio file
  * and optional configuration fields.
+ * - Uses 'json' format when timestamps are disabled (reduces token usage)
+ * - Uses 'verbose_json' format when timestamps are enabled (segment or word)
  *
  * @param options - Configuration for the multipart form data
  * @returns ArrayBuffer containing the complete multipart form data
  */
 export function buildMultipartFormData(options: MultipartFormDataOptions): ArrayBuffer {
-    const { boundary, audioBuffer, fileName, includeTimestamps, language, additionalFields = [] } = options;
+    const { boundary, audioBuffer, fileName, timestampGranularity, language, additionalFields = [] } = options;
     const parts: (string | Buffer)[] = [];
 
     // File field
@@ -115,8 +137,8 @@ export function buildMultipartFormData(options: MultipartFormDataOptions): Array
         parts.push(`${field.value}\r\n`);
     }
 
-    // Response format
-    const responseFormat = includeTimestamps ? 'verbose_json' : 'json';
+    // Use 'json' when timestamps disabled, 'verbose_json' when enabled
+    const responseFormat = timestampGranularity === 'disabled' ? 'json' : 'verbose_json';
     parts.push(`--${boundary}\r\n`);
     parts.push(`Content-Disposition: form-data; name="response_format"\r\n\r\n`);
     parts.push(`${responseFormat}\r\n`);
@@ -128,11 +150,11 @@ export function buildMultipartFormData(options: MultipartFormDataOptions): Array
         parts.push(`${language}\r\n`);
     }
 
-    // Timestamp granularities (for verbose_json)
-    if (includeTimestamps) {
+    // Only add timestamp_granularities[] when timestamps are enabled
+    if (timestampGranularity !== 'disabled') {
         parts.push(`--${boundary}\r\n`);
         parts.push(`Content-Disposition: form-data; name="timestamp_granularities[]"\r\n\r\n`);
-        parts.push(`segment\r\n`);
+        parts.push(`${timestampGranularity}\r\n`);
     }
 
     parts.push(`--${boundary}--\r\n`);
@@ -178,7 +200,7 @@ export function extractFileName(filePath: string, defaultName = 'audio.mp3'): st
  * @throws Error if the audio file cannot be read
  */
 export function prepareAudioFormData(options: PrepareAudioFormDataOptions): PreparedAudioFormData {
-    const { audioFilePath, includeTimestamps, language, additionalFields } = options;
+    const { audioFilePath, timestampGranularity, language, additionalFields } = options;
 
     validateAudioFile(audioFilePath);
 
@@ -190,7 +212,7 @@ export function prepareAudioFormData(options: PrepareAudioFormDataOptions): Prep
         boundary,
         audioBuffer,
         fileName,
-        includeTimestamps,
+        timestampGranularity,
         language,
         additionalFields,
     });
diff --git a/src/processing/index.ts b/src/processing/index.ts
index aa1d7d7..b488143 100644
--- a/src/processing/index.ts
+++ b/src/processing/index.ts
@@ -10,6 +10,8 @@ export {
 
 export type {
     TranscriptionApiResponse,
+    TranscriptionSegment,
+    TranscriptionWord,
     TranscriptionChunk,
     TranscriptionResult,
     FormField,
diff --git a/src/processing/workflow-chaining.ts b/src/processing/workflow-chaining.ts
index f972648..c660f9b 100644
--- a/src/processing/workflow-chaining.ts
+++ b/src/processing/workflow-chaining.ts
@@ -1,5 +1,6 @@
 import { App, MarkdownView } from 'obsidian';
-import { WorkflowConfig, AIToolboxSettings } from '../settings';
+import { WorkflowConfig, AIToolboxSettings, TimestampGranularity } from '../settings';
+import { TranscriptionResult, TranscriptionChunk } from '../providers';
 
 /**
  * Result from executing a workflow, used for chaining
@@ -121,26 +122,64 @@ export function createChatWorkflowTokens(
 }
 
 /**
- * Create transcription workflow tokens from execution data
+ * Format a timestamp in seconds to MM:SS or HH:MM:SS format.
+ */
+export function formatTimestamp(seconds: number): string {
+    const hours = Math.floor(seconds / 3600);
+    const minutes = Math.floor((seconds % 3600) / 60);
+    const secs = Math.floor(seconds % 60);
+
+    if (hours > 0) {
+        return `${hours.toString().padStart(2, '0')}:${minutes.toString().padStart(2, '0')}:${secs.toString().padStart(2, '0')}`;
+    }
+    return `${minutes.toString().padStart(2, '0')}:${secs.toString().padStart(2, '0')}`;
+}
+
+/**
+ * Format transcription chunks with timestamps.
+ * Each segment is prefixed with its start time in [MM:SS] format.
+ */
+export function formatTranscriptionWithTimestamps(chunks: TranscriptionChunk[]): string {
+    if (!chunks || chunks.length === 0) {
+        return '';
+    }
+
+    return chunks
+        .map(chunk => `[${formatTimestamp(chunk.timestamp[0])}] ${chunk.text}`)
+        .join('\n');
+}
+
+/**
+ * Create transcription workflow tokens from execution data.
+ * Generates plain text transcription and optionally timestamped version.
+ * The transcriptionWithTimestamps token is only included when granularity is not 'disabled'.
  */
 export function createTranscriptionWorkflowTokens(
-    transcriptionText: string,
+    transcriptionResult: TranscriptionResult,
     metadata?: {
         title?: string;
         uploader?: string;
         sourceUrl?: string;
         description?: string;
         tags?: string[];
-    }
+    },
+    timestampGranularity?: TimestampGranularity
 ): Record<string, string> {
-    return {
-        transcription: transcriptionText,
+    const tokens: Record<string, string> = {
+        transcription: transcriptionResult.text,
         title: metadata?.title ?? '',
         author: metadata?.uploader ?? '',
         sourceUrl: metadata?.sourceUrl ?? '',
         description: metadata?.description ?? '',
         tags: metadata?.tags?.join(', ') ?? ''
     };
+
+    // Only include timestamped transcription when timestamps are enabled
+    if (timestampGranularity !== 'disabled') {
+        tokens.transcriptionWithTimestamps = formatTranscriptionWithTimestamps(transcriptionResult.chunks);
+    }
+
+    return tokens;
 }
 
 /**
diff --git a/src/processing/workflow-executor.ts b/src/processing/workflow-executor.ts
index ff37384..7c03290 100644
--- a/src/processing/workflow-executor.ts
+++ b/src/processing/workflow-executor.ts
@@ -27,7 +27,8 @@ import {
     getDependencyWorkflowIds,
     gatherContextValues,
     replaceContextTokens,
-    hasContextTokens
+    hasContextTokens,
+    formatTranscriptionWithTimestamps
 } from './workflow-chaining';
 
 /**
@@ -61,35 +62,6 @@ function createInputHandler(sourceType: TranscriptionSourceType): InputHandler {
 	}
 }
 
-/**
- * Format a timestamp in seconds to MM:SS or HH:MM:SS format.
- */
-function formatTimestamp(seconds: number): string {
-	const hours = Math.floor(seconds / 3600);
-	const minutes = Math.floor((seconds % 3600) / 60);
-	const secs = Math.floor(seconds % 60);
-
-	if (hours > 0) {
-		return `${hours.toString().padStart(2, '0')}:${minutes.toString().padStart(2, '0')}:${secs.toString().padStart(2, '0')}`;
-	}
-	return `${minutes.toString().padStart(2, '0')}:${secs.toString().padStart(2, '0')}`;
-}
-
-/**
- * Format a transcription result, optionally including timestamps.
- * When timestamps are included and chunks are available, formats each segment
- * with its start time prefix.
- */
-function formatTranscriptionResult(result: TranscriptionResult, includeTimestamps: boolean): string {
-	if (!includeTimestamps || !result.chunks || result.chunks.length === 0) {
-		return result.text;
-	}
-
-	return result.chunks
-		.map(chunk => `[${formatTimestamp(chunk.timestamp[0])}] ${chunk.text}`)
-		.join('\n');
-}
-
 /**
  * Generate a note title for transcription output based on input source.
  * - TikTok: "TikTok by <Author> - <Timestamp>"
@@ -466,12 +438,15 @@ async function executeTranscriptionWorkflow(
 		new Notice(`Transcribing audio...`);
 
 		const transcriptionOptions: TranscriptionOptions = {
-			includeTimestamps: workflow.includeTimestamps ?? true,
+			timestampGranularity: workflow.timestampGranularity ?? 'disabled',
 			language: workflow.language || undefined
 		};
 
 		const transcriptionResult = await provider.transcribeAudio(inputResult.audioFilePath, transcriptionOptions);
-		const formattedText = formatTranscriptionResult(transcriptionResult, workflow.includeTimestamps ?? true);
+		// Use timestamped version for display output if available, otherwise plain text
+		const formattedText = transcriptionResult.chunks.length > 0
+			? formatTranscriptionWithTimestamps(transcriptionResult.chunks)
+			: transcriptionResult.text;
 		const noteTitle = generateTranscriptionNoteTitle(inputResult, workflow.name);
 
 		const outputType = workflow.outputType || 'new-note';
@@ -526,18 +501,18 @@ async function executeTranscriptionWorkflowInternal(
 			return { ...baseResult, error: 'No input provided or cancelled' };
 		}
 
+		const timestampGranularity = workflow.timestampGranularity ?? 'disabled';
 		const transcriptionOptions: TranscriptionOptions = {
-			includeTimestamps: workflow.includeTimestamps ?? true,
+			timestampGranularity,
 			language: workflow.language || undefined
 		};
 
 		const transcriptionResult = await provider.transcribeAudio(inputResult.audioFilePath, transcriptionOptions);
-		const formattedText = formatTranscriptionResult(transcriptionResult, workflow.includeTimestamps ?? true);
 
 		return {
 			...baseResult,
 			success: true,
-			tokens: createTranscriptionWorkflowTokens(formattedText, inputResult.metadata)
+			tokens: createTranscriptionWorkflowTokens(transcriptionResult, inputResult.metadata, timestampGranularity)
 		};
 	} catch (error) {
 		const errorMessage = error instanceof Error ? error.message : String(error);
diff --git a/src/providers/base-provider.ts b/src/providers/base-provider.ts
index 0d5acf6..35aa9bf 100644
--- a/src/providers/base-provider.ts
+++ b/src/providers/base-provider.ts
@@ -1,6 +1,6 @@
 import { Notice, requestUrl } from 'obsidian';
-import { ModelProvider, ModelProviderConfig, TranscriptionOptions, TranscriptionResult, ChatMessage, ChatOptions, ChatResult, TestAudioData } from './types';
-import { AIProviderType } from '../settings';
+import { ModelProvider, ModelProviderConfig, TranscriptionOptions, TranscriptionResult, ChatMessage, ChatOptions, ChatResult, TestAudioData, TranscriptionWord } from './types';
+import { AIProviderType, TimestampGranularity } from '../settings';
 import { prepareAudioFormData, TranscriptionApiResponse, FormField, buildMultipartFormData, generateFormBoundary } from '../processing/audio-processor';
 
 /**
@@ -98,9 +98,11 @@ export abstract class BaseProvider implements ModelProvider {
 		try {
 			new Notice(`Transcribing audio with ${this.getProviderDisplayName()}`);
 
+			const timestampGranularity = options.timestampGranularity ?? 'segment';
+
 			const { boundary, formData } = prepareAudioFormData({
 				audioFilePath,
-				includeTimestamps: options.includeTimestamps || false,
+				timestampGranularity,
 				language: options.language,
 				additionalFields: this.getAdditionalFormFields(),
 			});
@@ -108,7 +110,7 @@ export abstract class BaseProvider implements ModelProvider {
 			const result = await this.sendTranscriptionRequest(boundary, formData);
 			new Notice('Transcription complete!');
 
-			return this.parseTranscriptionResponse(result, audioFilePath, options.includeTimestamps || false);
+			return this.parseTranscriptionResponse(result, audioFilePath, timestampGranularity);
 		} catch (error) {
 			console.error('Transcription error:', error);
 			const errorMessage = error instanceof Error ? error.message : String(error);
@@ -125,7 +127,7 @@ export abstract class BaseProvider implements ModelProvider {
 			boundary,
 			audioBuffer: testAudio.audioBuffer,
 			fileName: testAudio.fileName,
-			includeTimestamps: false,
+			timestampGranularity: 'segment',
 			additionalFields: this.getAdditionalFormFields(),
 		});
 
@@ -220,21 +222,36 @@ export abstract class BaseProvider implements ModelProvider {
 		return result;
 	}
 
+	/**
+	 * Parse transcription API response into a TranscriptionResult.
+	 * - When granularity is 'disabled': returns empty chunks (API uses 'json' format with no timestamps)
+	 * - When granularity is 'segment': extracts segment-level timestamps
+	 * - When granularity is 'word': extracts both segment and word-level timestamps
+	 */
 	protected parseTranscriptionResponse(
 		response: TranscriptionApiResponse,
 		audioFilePath: string,
-		includeTimestamps: boolean
+		granularity: TimestampGranularity
 	): TranscriptionResult {
+		// Extract segment-level chunks (empty array when granularity is 'disabled')
+		const chunks = (response.segments ?? []).map(segment => ({
+			text: segment.text.trim(),
+			timestamp: [segment.start, segment.end] as [number, number],
+		}));
+
 		const result: TranscriptionResult = {
 			text: response.text,
+			chunks,
 			audioFilePath,
 		};
 
-		if (includeTimestamps && response.segments) {
-			result.chunks = response.segments.map(segment => ({
-				text: segment.text.trim(),
-				timestamp: [segment.start, segment.end] as [number, number],
-			}));
+		// Include word-level timestamps if granularity is 'word' and words are available
+		if (granularity === 'word' && response.words) {
+			result.words = response.words.map(word => ({
+				word: word.word,
+				start: word.start,
+				end: word.end,
+			} as TranscriptionWord));
 		}
 
 		return result;
diff --git a/src/providers/types.ts b/src/providers/types.ts
index ec23f59..a1b66e6 100644
--- a/src/providers/types.ts
+++ b/src/providers/types.ts
@@ -1,19 +1,19 @@
-import { AIProviderType } from '../settings';
-import { TestAudioData } from '../processing/audio-processor';
+import { AIProviderType, TimestampGranularity } from '../settings';
+import { TestAudioData, TranscriptionWord, TranscriptionSegment } from '../processing/audio-processor';
 
-// Re-export TestAudioData for consumers of this module
-export type { TestAudioData };
+// Re-export types for consumers of this module
+export type { TestAudioData, TranscriptionWord, TranscriptionSegment };
 
 /**
  * Options for audio transcription
  */
 export interface TranscriptionOptions {
-	includeTimestamps?: boolean;
+	timestampGranularity?: TimestampGranularity;
 	language?: string;
 }
 
 /**
- * Individual chunk with timestamps from transcription
+ * Individual chunk with timestamps from transcription (segment-level)
  */
 export interface TranscriptionChunk {
 	text: string;
@@ -21,11 +21,17 @@ export interface TranscriptionChunk {
 }
 
 /**
- * Result from audio transcription
+ * Result from audio transcription.
+ * Always includes both plain text and timestamped chunks.
  */
 export interface TranscriptionResult {
+	/** The complete transcription text */
 	text: string;
-	chunks?: TranscriptionChunk[];
+	/** Timestamped segments/chunks from the transcription */
+	chunks: TranscriptionChunk[];
+	/** Word-level timestamps (only populated when granularity is 'word') */
+	words?: TranscriptionWord[];
+	/** Path to the source audio file */
 	audioFilePath: string;
 }
 
diff --git a/src/settings/additional-settings.ts b/src/settings/additional-settings.ts
index bf53c52..9e413f8 100644
--- a/src/settings/additional-settings.ts
+++ b/src/settings/additional-settings.ts
@@ -6,6 +6,7 @@ import AIToolboxPlugin from "../main";
  */
 export interface AdditionalSettingsCallbacks {
 	refresh: () => void;
+	isAdvancedVisible: () => boolean;
 }
 
 /**
@@ -14,27 +15,9 @@ export interface AdditionalSettingsCallbacks {
 export function displayAdditionalSettings(
 	containerEl: HTMLElement,
 	plugin: AIToolboxPlugin,
-	_callbacks: AdditionalSettingsCallbacks
+	callbacks: AdditionalSettingsCallbacks
 ): void {
-	let showAdvanced = false;
-
-	const advancedToggleSetting = new Setting(containerEl)
-		.addToggle(toggle => toggle
-			.setValue(showAdvanced)
-			.onChange((value) => {
-				showAdvanced = value;
-				ytdlpPathSetting.settingEl.toggleClass('additional-settings-advanced-hidden', !value);
-				ffmpegPathSetting.settingEl.toggleClass('additional-settings-advanced-hidden', !value);
-				advancedLabel.toggleClass('additional-settings-advanced-label-active', value);
-				ytdlpPathSetting.nameEl.addClass('additional-settings-advanced-name');
-				ffmpegPathSetting.nameEl.addClass('additional-settings-advanced-name');
-			}));
-	advancedToggleSetting.settingEl.addClass('additional-settings-toggle');
-	const advancedLabel = advancedToggleSetting.controlEl.createSpan({
-		text: 'Show advanced settings',
-		cls: 'additional-settings-toggle-label'
-	});
-	advancedToggleSetting.controlEl.prepend(advancedLabel);
+	const showAdvanced = callbacks.isAdvancedVisible();
 
 	// yt-dlp section header
 	const ytdlpHeading = new Setting(containerEl)
@@ -67,7 +50,10 @@ export function displayAdditionalSettings(
 				plugin.settings.ytdlpLocation = value;
 				await plugin.saveSettings();
 			}));
-	ytdlpPathSetting.settingEl.addClass('additional-settings-advanced-hidden');
+	ytdlpPathSetting.settingEl.toggleClass('settings-advanced-hidden', !showAdvanced);
+	if (showAdvanced) {
+		ytdlpPathSetting.nameEl.addClass('settings-advanced-name');
+	}
 
 	const ffmpegPathSetting = new Setting(containerEl)
 		.setName('FFmpeg path') // eslint-disable-line obsidianmd/ui/sentence-case -- proper noun
@@ -78,6 +64,9 @@ export function displayAdditionalSettings(
 				plugin.settings.ffmpegLocation = value;
 				await plugin.saveSettings();
 			}));
-	ffmpegPathSetting.settingEl.addClass('additional-settings-advanced-hidden');
+	ffmpegPathSetting.settingEl.toggleClass('settings-advanced-hidden', !showAdvanced);
+	if (showAdvanced) {
+		ffmpegPathSetting.nameEl.addClass('settings-advanced-name');
+	}
 }
 
diff --git a/src/settings/index.ts b/src/settings/index.ts
index 10c8085..77e48a2 100644
--- a/src/settings/index.ts
+++ b/src/settings/index.ts
@@ -1,4 +1,4 @@
-import { App, PluginSettingTab } from "obsidian";
+import { App, PluginSettingTab, Setting } from "obsidian";
 import AIToolboxPlugin from "../main";
 import { SettingsTabType, ExpandOnNextRenderState } from "./types";
 import { displayProvidersSettings, ProviderSettingsCallbacks } from "./providers";
@@ -19,6 +19,7 @@ export type {
 	TranscriptionMediaType,
 	TranscriptionSourceType,
 	TranscriptionContextConfig,
+	TimestampGranularity,
 	ChatContextType,
 	ChatContextConfig,
 	AIToolboxSettings,
@@ -31,6 +32,8 @@ export class AIToolboxSettingTab extends PluginSettingTab {
 	private activeTab: SettingsTabType = 'providers';
 	// Track IDs that should start expanded on next render (cleared after use)
 	private expandOnNextRender: ExpandOnNextRenderState = {};
+	// Global advanced settings visibility state
+	private showAdvancedSettings = false;
 
 	constructor(app: App, plugin: AIToolboxPlugin) {
 		super(app, plugin);
@@ -43,6 +46,27 @@ export class AIToolboxSettingTab extends PluginSettingTab {
 		containerEl.empty();
 
 		const tabContainer = containerEl.createDiv('settings-tab-container');
+
+		// Global advanced settings toggle - positioned above tab header
+		const advancedToggleSetting = new Setting(tabContainer)
+			.addToggle(toggle => toggle
+				.setValue(this.showAdvancedSettings)
+				.onChange((value) => {
+					this.showAdvancedSettings = value;
+					advancedLabel.toggleClass('settings-advanced-toggle-label-active', value);
+					this.display();
+				}));
+		advancedToggleSetting.settingEl.addClass('settings-advanced-toggle');
+		const advancedLabel = advancedToggleSetting.controlEl.createSpan({
+			text: 'Show advanced settings',
+			cls: 'settings-advanced-toggle-label'
+		});
+		advancedToggleSetting.controlEl.prepend(advancedLabel);
+		// Apply active state on initial render if already enabled
+		if (this.showAdvancedSettings) {
+			advancedLabel.addClass('settings-advanced-toggle-label-active');
+		}
+
 		const tabHeader = tabContainer.createDiv('settings-tab-header');
 		const tabContent = tabContainer.createDiv('settings-tab-content');
 
@@ -100,14 +124,16 @@ export class AIToolboxSettingTab extends PluginSettingTab {
 		const callbacks: WorkflowSettingsCallbacks = {
 			getExpandState: () => this.expandOnNextRender,
 			setExpandState: (state) => { this.expandOnNextRender = state; },
-			refresh: () => this.display()
+			refresh: () => this.display(),
+			isAdvancedVisible: () => this.showAdvancedSettings
 		};
 		displayWorkflowsSettings(containerEl, this.plugin, callbacks);
 	}
 
 	private displaySettingsTab(containerEl: HTMLElement): void {
 		const callbacks: AdditionalSettingsCallbacks = {
-			refresh: () => this.display()
+			refresh: () => this.display(),
+			isAdvancedVisible: () => this.showAdvancedSettings
 		};
 		displayAdditionalSettings(containerEl, this.plugin, callbacks);
 	}
diff --git a/src/settings/types.ts b/src/settings/types.ts
index e881379..77fcdc1 100644
--- a/src/settings/types.ts
+++ b/src/settings/types.ts
@@ -65,6 +65,14 @@ export type TranscriptionMediaType = 'video' | 'audio';
  */
 export type TranscriptionSourceType = 'select-file-from-vault' | 'url-from-clipboard' | 'url-from-selection';
 
+/**
+ * Timestamp granularity for transcription output.
+ * - 'disabled': No timestamps (uses simpler JSON format, reduces token usage)
+ * - 'segment': Timestamps at sentence/phrase level
+ * - 'word': Timestamps at individual word level
+ */
+export type TimestampGranularity = 'disabled' | 'segment' | 'word';
+
 /**
  * Context configuration for transcription workflows
  */
@@ -119,7 +127,7 @@ export interface WorkflowConfig {
 	workflowContexts?: WorkflowContextConfig[];
 	// Transcription-specific settings (optional for backward compatibility)
 	language?: string;
-	includeTimestamps?: boolean;
+	timestampGranularity?: TimestampGranularity;
 	transcriptionContext?: TranscriptionContextConfig;
 }
 
@@ -141,7 +149,7 @@ export const DEFAULT_WORKFLOW_CONFIG: Omit<WorkflowConfig, 'id'> = {
 	contexts: [],
 	workflowContexts: [],
 	language: '',
-	includeTimestamps: true
+	timestampGranularity: 'disabled'
 };
 
 export interface AIToolboxSettings {
diff --git a/src/settings/workflows.ts b/src/settings/workflows.ts
index 8e60ca0..1903cd7 100644
--- a/src/settings/workflows.ts
+++ b/src/settings/workflows.ts
@@ -7,6 +7,7 @@ import {
 	PromptSourceType,
 	TranscriptionMediaType,
 	TranscriptionSourceType,
+	TimestampGranularity,
 	ChatContextType,
 	ExpandOnNextRenderState,
 	generateId,
@@ -30,6 +31,7 @@ export interface WorkflowSettingsCallbacks {
 	getExpandState: () => ExpandOnNextRenderState;
 	setExpandState: (state: ExpandOnNextRenderState) => void;
 	refresh: () => void;
+	isAdvancedVisible: () => boolean;
 }
 
 /**
@@ -99,6 +101,8 @@ function displayWorkflowSettings(
 	const workflowType = workflow.type || 'chat';
 	const icon = workflowType === 'chat' ? 'message-circle' : 'audio-lines';
 
+	const showAdvanced = callbacks.isAdvancedVisible();
+
 	const { contentContainer, updateTitle, isExpanded } = createCollapsibleSection({
 		containerEl,
 		title: workflow.name || 'Unnamed workflow',
@@ -108,7 +112,7 @@ function displayWorkflowSettings(
 		startExpanded: shouldExpand,
 		isHeading: true,
 		icon,
-		secondaryText: workflow.id,
+		secondaryText: showAdvanced ? workflow.id : undefined,
 		onDelete: async () => {
 			const index = plugin.settings.workflows.findIndex(w => w.id === workflow.id);
 			if (index !== -1) {
@@ -435,8 +439,9 @@ function displayTranscriptionWorkflowSettings(
 				await plugin.saveSettings();
 			}));
 
-	// Language setting
-	new Setting(contentContainer)
+	// Language setting (advanced)
+	const showAdvanced = callbacks.isAdvancedVisible();
+	const languageSetting = new Setting(contentContainer)
 		.setName('Language')
 		.setDesc('Optional language code for transcription (e.g., "en", "es", "fr"). Leave empty for auto-detection.')
 		.addText(text => text
@@ -446,6 +451,33 @@ function displayTranscriptionWorkflowSettings(
 				workflow.language = value;
 				await plugin.saveSettings();
 			}));
+	languageSetting.settingEl.toggleClass('settings-advanced-hidden', !showAdvanced);
+	if (showAdvanced) {
+		languageSetting.nameEl.addClass('settings-advanced-name');
+	}
+
+	// Timestamp granularity dropdown (advanced)
+	const granularityOptions: Record<TimestampGranularity, string> = {
+		'disabled': 'Disabled (no timestamps)',
+		'segment': 'Segment (sentence/phrase level)',
+		'word': 'Word (individual word level)'
+	};
+	const granularitySetting = new Setting(contentContainer)
+		.setName('Timestamp granularity')
+		.setDesc('Level of detail for timestamps. Disabling reduces token usage.')
+		.addDropdown(dropdown => dropdown
+			.addOptions(granularityOptions)
+			.setValue(workflow.timestampGranularity ?? 'disabled')
+			.onChange(async (value) => {
+				workflow.timestampGranularity = value as TimestampGranularity;
+				await plugin.saveSettings();
+				callbacks.setExpandState({ workflowId: workflow.id });
+				callbacks.refresh();
+			}));
+	granularitySetting.settingEl.toggleClass('settings-advanced-hidden', !showAdvanced);
+	if (showAdvanced) {
+		granularitySetting.nameEl.addClass('settings-advanced-name');
+	}
 
 	// Make available as input to other workflows toggle
 	new Setting(contentContainer)
@@ -497,17 +529,6 @@ function displayTranscriptionWorkflowSettings(
 					callbacks.refresh();
 				}));
 
-		// Include timestamps toggle
-		new Setting(contentContainer)
-			.setName('Include timestamps')
-			.setDesc('Include timestamps in the transcription output')
-			.addToggle(toggle => toggle
-				.setValue(workflow.includeTimestamps ?? true)
-				.onChange(async (value) => {
-					workflow.includeTimestamps = value;
-					await plugin.saveSettings();
-				}));
-
 		// Output folder (only show if output type is new-note)
 		if (workflow.outputType === 'new-note') {
 			createPathPicker({
@@ -719,6 +740,8 @@ interface TokenGroup {
 	workflowId?: string;
 	/** For workflow groups: the source workflow type */
 	workflowType?: WorkflowType;
+	/** For transcription workflows: the timestamp granularity setting */
+	timestampGranularity?: TimestampGranularity;
 }
 
 /**
@@ -789,16 +812,19 @@ function displayAvailableTokensSection(
 		if (!sourceWorkflow) continue;
 
 		const workflowType = sourceWorkflow.type || 'chat';
+		const timestampGranularity = sourceWorkflow.timestampGranularity;
 		const workflowTokens = getWorkflowContextTokens(
 			sourceWorkflow.id,
-			workflowType
+			workflowType,
+			{ timestampGranularity }
 		);
 		if (workflowTokens.length > 0) {
 			tokenGroups.push({
 				name: sourceWorkflow.name || 'Unnamed workflow',
 				tokens: workflowTokens,
 				workflowId: sourceWorkflow.id,
-				workflowType: workflowType
+				workflowType: workflowType,
+				timestampGranularity: timestampGranularity
 			});
 			totalTokenCount += workflowTokens.length;
 		}
@@ -843,9 +869,10 @@ function displayAvailableTokensSection(
 			groupHeader.setAttribute('title', 'Click to copy all tokens as template');
 			const workflowId = group.workflowId;
 			const workflowType = group.workflowType;
+			const timestampGranularity = group.timestampGranularity;
 			groupHeader.addEventListener('click', (e) => {
 				e.stopPropagation();
-				const template = generateWorkflowTokenTemplate(workflowId, workflowType);
+				const template = generateWorkflowTokenTemplate(workflowId, workflowType, { timestampGranularity });
 				void navigator.clipboard.writeText(template).then(() => {
 					new Notice('Copied token template to clipboard');
 				});
diff --git a/src/tokens/types.ts b/src/tokens/types.ts
index c302971..f2c2947 100644
--- a/src/tokens/types.ts
+++ b/src/tokens/types.ts
@@ -1,4 +1,4 @@
-import { WorkflowType } from "../settings/types";
+import { WorkflowType, TimestampGranularity } from "../settings/types";
 
 /**
  * Token definition for display in the settings UI
@@ -36,8 +36,10 @@ export interface TranscriptionWorkflowTokens {
 	author: string;
 	/** The title of the video */
 	title: string;
-	/** The full transcription text */
+	/** The full transcription text (plain text without timestamps) */
 	transcription: string;
+	/** The transcription with timestamps (formatted with [MM:SS] prefixes) */
+	transcriptionWithTimestamps: string;
 	/** The original video URL */
 	sourceUrl: string;
 	/** The video description */
@@ -63,7 +65,8 @@ export const TRANSCRIPTION_WORKFLOW_TOKENS: TokenDefinition[] = [
 	{ name: 'sourceUrl', description: 'The original video URL' },
 	{ name: 'description', description: 'The video description' },
 	{ name: 'tags', description: 'The video tags (comma-separated)' },
-	{ name: 'transcription', description: 'The full transcription text' }
+	{ name: 'transcription', description: 'The plain transcription text (no timestamps)' },
+	{ name: 'transcriptionWithTimestamps', description: 'The transcription with [MM:SS] timestamps' }
 ];
 
 /**
@@ -77,16 +80,37 @@ const TOKEN_LABELS: Record<string, string> = {
 	sourceUrl: 'Source URL',
 	description: 'Description',
 	tags: 'Tags',
-	transcription: 'Transcription'
+	transcription: 'Transcription',
+	transcriptionWithTimestamps: 'Transcription (with timestamps)'
 };
 
 /**
- * Get token definitions for a workflow type
+ * Options for getting token definitions
  */
-export function getTokenDefinitionsForType(type: WorkflowType): TokenDefinition[] {
-	return type === 'transcription'
-		? TRANSCRIPTION_WORKFLOW_TOKENS
-		: CHAT_WORKFLOW_TOKENS;
+export interface TokenDefinitionOptions {
+	/** For transcription workflows, the timestamp granularity setting */
+	timestampGranularity?: TimestampGranularity;
+}
+
+/**
+ * Get token definitions for a workflow type.
+ * For transcription workflows, the transcriptionWithTimestamps token is excluded
+ * when timestampGranularity is 'disabled'.
+ */
+export function getTokenDefinitionsForType(
+	type: WorkflowType,
+	options?: TokenDefinitionOptions
+): TokenDefinition[] {
+	if (type === 'transcription') {
+		// Filter out transcriptionWithTimestamps when granularity is disabled
+		if (options?.timestampGranularity === 'disabled') {
+			return TRANSCRIPTION_WORKFLOW_TOKENS.filter(
+				token => token.name !== 'transcriptionWithTimestamps'
+			);
+		}
+		return TRANSCRIPTION_WORKFLOW_TOKENS;
+	}
+	return CHAT_WORKFLOW_TOKENS;
 }
 
 /**
@@ -95,12 +119,14 @@ export function getTokenDefinitionsForType(type: WorkflowType): TokenDefinition[
  *
  * @param workflowId - The ID of the workflow
  * @param workflowType - The type of the workflow (chat or transcription)
+ * @param options - Optional settings like timestampGranularity
  */
 export function getWorkflowContextTokens(
 	workflowId: string,
-	workflowType: WorkflowType
+	workflowType: WorkflowType,
+	options?: TokenDefinitionOptions
 ): TokenDefinition[] {
-	const baseTokens = getTokenDefinitionsForType(workflowType);
+	const baseTokens = getTokenDefinitionsForType(workflowType, options);
 
 	return baseTokens.map(token => ({
 		name: `${workflowId}.${token.name}`,
@@ -114,12 +140,14 @@ export function getWorkflowContextTokens(
  *
  * @param workflowId - The ID of the workflow
  * @param workflowType - The type of the workflow (chat or transcription)
+ * @param options - Optional settings like timestampGranularity
  */
 export function generateWorkflowTokenTemplate(
 	workflowId: string,
-	workflowType: WorkflowType
+	workflowType: WorkflowType,
+	options?: TokenDefinitionOptions
 ): string {
-	const tokens = getTokenDefinitionsForType(workflowType);
+	const tokens = getTokenDefinitionsForType(workflowType, options);
 
 	return tokens
 		.map(token => {
diff --git a/styles.css b/styles.css
index d981e3a..8912930 100644
--- a/styles.css
+++ b/styles.css
@@ -601,32 +601,37 @@ button.mod-success:hover {
 	color: var(--text-on-accent);
 }
 
-/* Additional settings styles */
-.additional-settings-toggle {
+/* Global advanced settings toggle (above tab header) */
+.settings-advanced-toggle {
 	justify-content: flex-end;
+	margin-bottom: 8px;
+	border-bottom: none;
 }
 
-.additional-settings-toggle .setting-item-name {
+.settings-advanced-toggle .setting-item-name {
 	display: none;
 }
 
-.additional-settings-toggle-label {
+.settings-advanced-toggle-label {
 	font-size: 0.85em;
 	margin-right: 8px;
 }
 
-.additional-settings-heading .setting-item-name {
-	font-size: 1.17em;
+.settings-advanced-toggle-label-active {
+	color: var(--text-warning);
 }
 
-.additional-settings-advanced-hidden {
+/* Settings hidden when advanced toggle is off */
+.settings-advanced-hidden {
 	display: none;
 }
 
-.additional-settings-advanced-label-active {
+/* Advanced setting name styling when visible */
+.settings-advanced-name {
 	color: var(--text-warning);
 }
 
-.additional-settings-advanced-name {
-	color: var(--text-warning);
+/* Additional settings tab styles */
+.additional-settings-heading .setting-item-name {
+	font-size: 1.17em;
 }

From dcdaf122cc3b9b28def15848b44188e20b462f0f Mon Sep 17 00:00:00 2001
From: dalinicus <contact@dalinicus.com>
Date: Fri, 9 Jan 2026 21:02:55 -0600
Subject: [PATCH 2/7] change name to llm studio

---
 README.md                           | 12 ++++++------
 manifest.json                       |  8 ++++----
 src/handlers/input/types.ts         |  4 ++--
 src/main.ts                         | 10 +++++-----
 src/processing/video-processor.ts   | 10 +++++-----
 src/processing/workflow-chaining.ts |  6 +++---
 src/processing/workflow-executor.ts | 24 +++++++++++++++---------
 src/providers/provider-factory.ts   |  6 +++---
 src/settings/additional-settings.ts |  4 ++--
 src/settings/index.ts               | 10 +++++-----
 src/settings/providers.ts           |  8 ++++----
 src/settings/types.ts               |  4 ++--
 src/settings/workflows.ts           | 20 ++++++++++----------
 src/tokens/types.ts                 |  7 ++++---
 14 files changed, 70 insertions(+), 63 deletions(-)

diff --git a/README.md b/README.md
index 91b9b0d..b3fe3ae 100644
--- a/README.md
+++ b/README.md
@@ -1,6 +1,6 @@
-# AI Toolbox
+# LLM Toolbox
 
-A personal collection of AI tools to enhance the Obsidian.md workflow.
+A personal collection of LLM tools to enhance the Obsidian.md workflow.
 
 ## Features
 
@@ -37,7 +37,7 @@ pip install yt-dlp[curl_cffi]
 ## How to Use
 
 1. **Configure the plugin**:
-   - Go to **Settings → AI Toolbox**
+   - Go to **Settings → LLM Toolbox**
    - Enter your Azure OpenAI endpoint, API key, and deployment name
    - Configure optional settings (timestamps, language, output folder)
 
@@ -53,14 +53,14 @@ pip install yt-dlp[curl_cffi]
 
 1. Install the [BRAT plugin](https://github.com/TfTHacker/obsidian42-brat) if you haven't already
 2. Open **Settings → BRAT → Add Beta plugin**
-3. Enter this repository ID: `dalinicus/obsidian-ai-toolbox`
+3. Enter this repository ID: `dalinicus/obsidian-llm-toolbox`
 4. Click **Add Plugin**
 5. Enable the plugin in **Settings → Community plugins**
 
 ### Manual Installation
 
-1. Download the latest release from the [Releases page](https://github.com/dalinicus/obsidian-ai-toolbox/releases)
-2. Extract the files to your vault's plugins folder: `VaultFolder/.obsidian/plugins/ai-toolbox/`
+1. Download the latest release from the [Releases page](https://github.com/dalinicus/obsidian-llm-toolbox/releases)
+2. Extract the files to your vault's plugins folder: `VaultFolder/.obsidian/plugins/llm-toolbox/`
 3. Reload Obsidian
 4. Enable the plugin in **Settings → Community plugins**
 
diff --git a/manifest.json b/manifest.json
index de98265..ba922aa 100644
--- a/manifest.json
+++ b/manifest.json
@@ -1,9 +1,9 @@
 {
-	"id": "ai-toolbox",
-	"name": "AI Toolbox",
-	"version": "1.0.3",
+	"id": "llm-toolbox",
+	"name": "LLM Toolbox",
+	"version": "1.0.5",
 	"minAppVersion": "0.15.0",
-	"description": "Personal collection of AI tools to enhance the Obsidian.md workflow.",
+	"description": "Personal collection of LLM tools to enhance the Obsidian.md workflow.",
 	"author": "dalinicus",
 	"authorUrl": "https://github.com/dalinicus",
 	"fundingUrl": "https://buymeacoffee.com/dalinicus",
diff --git a/src/handlers/input/types.ts b/src/handlers/input/types.ts
index 6e0d6e7..deb19a0 100644
--- a/src/handlers/input/types.ts
+++ b/src/handlers/input/types.ts
@@ -1,5 +1,5 @@
 import { App } from 'obsidian';
-import { AIToolboxSettings, WorkflowConfig } from '../../settings';
+import { LLMToolboxSettings, WorkflowConfig } from '../../settings';
 import { VideoMetadata } from '../../processing';
 
 /**
@@ -22,7 +22,7 @@ export interface InputContext {
     /** The Obsidian App instance */
     app: App;
     /** Plugin settings */
-    settings: AIToolboxSettings;
+    settings: LLMToolboxSettings;
     /** The workflow configuration being executed */
     workflow: WorkflowConfig;
 }
diff --git a/src/main.ts b/src/main.ts
index 4e3b00c..c32deaf 100644
--- a/src/main.ts
+++ b/src/main.ts
@@ -1,10 +1,10 @@
 import { Notice, Plugin } from 'obsidian';
-import { DEFAULT_SETTINGS, AIToolboxSettings, AIToolboxSettingTab } from "./settings/index";
+import { DEFAULT_SETTINGS, LLMToolboxSettings, LLMToolboxSettingTab } from "./settings/index";
 import { WorkflowSuggesterModal } from "./components/workflow-suggester";
 import { executeWorkflow } from "./processing/workflow-executor";
 
-export default class AIToolboxPlugin extends Plugin {
-	settings: AIToolboxSettings;
+export default class LLMToolboxPlugin extends Plugin {
+	settings: LLMToolboxSettings;
 
 	async onload() {
 		await this.loadSettings();
@@ -17,7 +17,7 @@ export default class AIToolboxPlugin extends Plugin {
 		});
 
 		// This adds a settings tab so the user can configure various aspects of the plugin
-		this.addSettingTab(new AIToolboxSettingTab(this.app, this));
+		this.addSettingTab(new LLMToolboxSettingTab(this.app, this));
 	}
 
 	/**
@@ -38,7 +38,7 @@ export default class AIToolboxPlugin extends Plugin {
 	}
 
 	async loadSettings() {
-		this.settings = Object.assign({}, DEFAULT_SETTINGS, await this.loadData() as Partial<AIToolboxSettings>);
+		this.settings = Object.assign({}, DEFAULT_SETTINGS, await this.loadData() as Partial<LLMToolboxSettings>);
 	}
 
 	async saveSettings() {
diff --git a/src/processing/video-processor.ts b/src/processing/video-processor.ts
index 7e3e6f7..585c3eb 100644
--- a/src/processing/video-processor.ts
+++ b/src/processing/video-processor.ts
@@ -4,7 +4,7 @@ import * as os from 'os';
 import * as fs from 'fs';
 import { Buffer } from 'buffer';
 import { Notice } from 'obsidian';
-import { AIToolboxSettings } from '../settings';
+import { LLMToolboxSettings } from '../settings';
 import { videoPlatformRegistry, VideoMetadata } from './video-platforms';
 
 /**
@@ -198,9 +198,9 @@ export interface ExtractAudioResult {
 }
 
 /**
- * Converts AIToolboxSettings to VideoProcessorConfig.
+ * Converts LLMToolboxSettings to VideoProcessorConfig.
  */
-function settingsToProcessorConfig(settings: AIToolboxSettings): VideoProcessorConfig {
+function settingsToProcessorConfig(settings: LLMToolboxSettings): VideoProcessorConfig {
     return {
         ytdlpLocation: settings.ytdlpLocation,
         ffmpegLocation: settings.ffmpegLocation,
@@ -216,7 +216,7 @@ function settingsToProcessorConfig(settings: AIToolboxSettings): VideoProcessorC
  * Uses yt-dlp directly via child_process for audio extraction.
  * Requires yt-dlp and ffmpeg to be installed and available in PATH.
  */
-export async function extractAudioFromUrl(url: string, settings: AIToolboxSettings): Promise<ExtractAudioResult | null> {
+export async function extractAudioFromUrl(url: string, settings: LLMToolboxSettings): Promise<ExtractAudioResult | null> {
     try {
         if (!url || !url.trim()) {
             new Notice('URL is empty');
@@ -270,7 +270,7 @@ export async function extractAudioFromUrl(url: string, settings: AIToolboxSettin
  * Uses yt-dlp directly via child_process for audio extraction.
  * Requires yt-dlp and ffmpeg to be installed and available in PATH.
  */
-export async function extractAudioFromClipboard(settings: AIToolboxSettings): Promise<ExtractAudioResult | null> {
+export async function extractAudioFromClipboard(settings: LLMToolboxSettings): Promise<ExtractAudioResult | null> {
     try {
         const clipboardText = await navigator.clipboard.readText();
 
diff --git a/src/processing/workflow-chaining.ts b/src/processing/workflow-chaining.ts
index c660f9b..3d9189d 100644
--- a/src/processing/workflow-chaining.ts
+++ b/src/processing/workflow-chaining.ts
@@ -1,5 +1,5 @@
 import { App, MarkdownView } from 'obsidian';
-import { WorkflowConfig, AIToolboxSettings, TimestampGranularity } from '../settings';
+import { WorkflowConfig, LLMToolboxSettings, TimestampGranularity } from '../settings';
 import { TranscriptionResult, TranscriptionChunk } from '../providers';
 
 /**
@@ -30,7 +30,7 @@ export interface DependencyExecutionContext {
     /** The Obsidian App instance */
     app: App;
     /** Plugin settings */
-    settings: AIToolboxSettings;
+    settings: LLMToolboxSettings;
     /** Results from executed dependencies */
     results: WorkflowResultsMap;
     /** Workflow IDs currently in the execution stack (for circular detection) */
@@ -43,7 +43,7 @@ export interface DependencyExecutionContext {
  */
 export function detectCircularDependency(
     workflow: WorkflowConfig,
-    settings: AIToolboxSettings,
+    settings: LLMToolboxSettings,
     visited: Set<string> = new Set(),
     path: string[] = []
 ): string[] {
diff --git a/src/processing/workflow-executor.ts b/src/processing/workflow-executor.ts
index 7c03290..e11e175 100644
--- a/src/processing/workflow-executor.ts
+++ b/src/processing/workflow-executor.ts
@@ -1,5 +1,5 @@
 import { App, Notice, TFile } from 'obsidian';
-import { WorkflowConfig, AIToolboxSettings, TranscriptionSourceType } from '../settings';
+import { WorkflowConfig, LLMToolboxSettings, TranscriptionSourceType } from '../settings';
 import { createWorkflowProvider, ChatMessage, TranscriptionOptions, TranscriptionResult } from '../providers';
 import { videoPlatformRegistry } from './video-platforms';
 import { generateFilenameTimestamp } from '../utils/date-utils';
@@ -130,7 +130,7 @@ async function getPromptText(app: App, workflow: WorkflowConfig): Promise<string
  */
 async function executeDependencies(
 	app: App,
-	settings: AIToolboxSettings,
+	settings: LLMToolboxSettings,
 	workflow: WorkflowConfig,
 	results: WorkflowResultsMap,
 	executionStack: Set<string>
@@ -194,7 +194,7 @@ async function executeDependencies(
  */
 export async function executeWorkflow(
 	app: App,
-	settings: AIToolboxSettings,
+	settings: LLMToolboxSettings,
 	workflow: WorkflowConfig
 ): Promise<void> {
 	// Check for circular dependencies before starting
@@ -242,7 +242,7 @@ export async function executeWorkflow(
  */
 async function executeWorkflowInternal(
 	app: App,
-	settings: AIToolboxSettings,
+	settings: LLMToolboxSettings,
 	workflow: WorkflowConfig,
 	dependencyResults: WorkflowResultsMap
 ): Promise<WorkflowExecutionResult> {
@@ -260,7 +260,7 @@ async function executeWorkflowInternal(
  */
 async function executeChatWorkflow(
 	app: App,
-	settings: AIToolboxSettings,
+	settings: LLMToolboxSettings,
 	workflow: WorkflowConfig,
 	dependencyResults?: WorkflowResultsMap
 ): Promise<void> {
@@ -337,7 +337,7 @@ async function executeChatWorkflow(
  */
 async function executeChatWorkflowInternal(
 	app: App,
-	settings: AIToolboxSettings,
+	settings: LLMToolboxSettings,
 	workflow: WorkflowConfig,
 	dependencyResults: WorkflowResultsMap
 ): Promise<WorkflowExecutionResult> {
@@ -405,7 +405,7 @@ async function executeChatWorkflowInternal(
  */
 async function executeTranscriptionWorkflow(
 	app: App,
-	settings: AIToolboxSettings,
+	settings: LLMToolboxSettings,
 	workflow: WorkflowConfig,
 	_dependencyResults?: WorkflowResultsMap
 ): Promise<void> {
@@ -467,7 +467,7 @@ async function executeTranscriptionWorkflow(
  */
 async function executeTranscriptionWorkflowInternal(
 	app: App,
-	settings: AIToolboxSettings,
+	settings: LLMToolboxSettings,
 	workflow: WorkflowConfig,
 	_dependencyResults: WorkflowResultsMap
 ): Promise<WorkflowExecutionResult> {
@@ -509,10 +509,16 @@ async function executeTranscriptionWorkflowInternal(
 
 		const transcriptionResult = await provider.transcribeAudio(inputResult.audioFilePath, transcriptionOptions);
 
+		// Combine metadata with sourceUrl for token creation
+		const tokenMetadata = {
+			...inputResult.metadata,
+			sourceUrl: inputResult.sourceUrl
+		};
+
 		return {
 			...baseResult,
 			success: true,
-			tokens: createTranscriptionWorkflowTokens(transcriptionResult, inputResult.metadata, timestampGranularity)
+			tokens: createTranscriptionWorkflowTokens(transcriptionResult, tokenMetadata, timestampGranularity)
 		};
 	} catch (error) {
 		const errorMessage = error instanceof Error ? error.message : String(error);
diff --git a/src/providers/provider-factory.ts b/src/providers/provider-factory.ts
index cbef66b..e0c0277 100644
--- a/src/providers/provider-factory.ts
+++ b/src/providers/provider-factory.ts
@@ -1,4 +1,4 @@
-import { AIToolboxSettings, AIProviderConfig, AIModelConfig, DEFAULT_OPENAI_ENDPOINT, WorkflowConfig, ProviderModelSelection } from '../settings/index';
+import { LLMToolboxSettings, AIProviderConfig, AIModelConfig, DEFAULT_OPENAI_ENDPOINT, WorkflowConfig, ProviderModelSelection } from '../settings/index';
 import { ModelProvider, ModelProviderConfig } from './types';
 import { AzureOpenAIModelProvider } from './azure-openai-provider';
 import { OpenAIModelProvider } from './openai-provider';
@@ -62,7 +62,7 @@ export function createModelProvider(config: ModelProviderConfig): ModelProvider
  * @returns A configured ModelProvider, or null if not found
  * @throws ProviderCreationError if the provider cannot be created
  */
-function createProviderFromSelection(settings: AIToolboxSettings, selection: ProviderModelSelection): ModelProvider | null {
+function createProviderFromSelection(settings: LLMToolboxSettings, selection: ProviderModelSelection): ModelProvider | null {
 	const provider = settings.providers.find(p => p.id === selection.providerId);
 	if (!provider) {
 		return null;
@@ -85,7 +85,7 @@ function createProviderFromSelection(settings: AIToolboxSettings, selection: Pro
  * @returns A configured ModelProvider for the workflow, or null if not configured
  * @throws ProviderCreationError if the provider cannot be created
  */
-export function createWorkflowProvider(settings: AIToolboxSettings, workflow: WorkflowConfig): ModelProvider | null {
+export function createWorkflowProvider(settings: LLMToolboxSettings, workflow: WorkflowConfig): ModelProvider | null {
 	if (!workflow.provider) {
 		return null;
 	}
diff --git a/src/settings/additional-settings.ts b/src/settings/additional-settings.ts
index 9e413f8..cc57536 100644
--- a/src/settings/additional-settings.ts
+++ b/src/settings/additional-settings.ts
@@ -1,5 +1,5 @@
 import { Setting } from "obsidian";
-import AIToolboxPlugin from "../main";
+import LLMToolboxPlugin from "../main";
 
 /**
  * Callbacks for the additional settings tab to communicate with the main settings tab
@@ -14,7 +14,7 @@ export interface AdditionalSettingsCallbacks {
  */
 export function displayAdditionalSettings(
 	containerEl: HTMLElement,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	callbacks: AdditionalSettingsCallbacks
 ): void {
 	const showAdvanced = callbacks.isAdvancedVisible();
diff --git a/src/settings/index.ts b/src/settings/index.ts
index 77e48a2..6ec9093 100644
--- a/src/settings/index.ts
+++ b/src/settings/index.ts
@@ -1,5 +1,5 @@
 import { App, PluginSettingTab, Setting } from "obsidian";
-import AIToolboxPlugin from "../main";
+import LLMToolboxPlugin from "../main";
 import { SettingsTabType, ExpandOnNextRenderState } from "./types";
 import { displayProvidersSettings, ProviderSettingsCallbacks } from "./providers";
 import { displayWorkflowsSettings, WorkflowSettingsCallbacks } from "./workflows";
@@ -22,20 +22,20 @@ export type {
 	TimestampGranularity,
 	ChatContextType,
 	ChatContextConfig,
-	AIToolboxSettings,
+	LLMToolboxSettings,
 	SettingsTabType,
 	ExpandOnNextRenderState
 } from "./types";
 
-export class AIToolboxSettingTab extends PluginSettingTab {
-	plugin: AIToolboxPlugin;
+export class LLMToolboxSettingTab extends PluginSettingTab {
+	plugin: LLMToolboxPlugin;
 	private activeTab: SettingsTabType = 'providers';
 	// Track IDs that should start expanded on next render (cleared after use)
 	private expandOnNextRender: ExpandOnNextRenderState = {};
 	// Global advanced settings visibility state
 	private showAdvancedSettings = false;
 
-	constructor(app: App, plugin: AIToolboxPlugin) {
+	constructor(app: App, plugin: LLMToolboxPlugin) {
 		super(app, plugin);
 		this.plugin = plugin;
 	}
diff --git a/src/settings/providers.ts b/src/settings/providers.ts
index 76b2233..30c9b3c 100644
--- a/src/settings/providers.ts
+++ b/src/settings/providers.ts
@@ -1,5 +1,5 @@
 import { Setting, setIcon, ButtonComponent, Notice } from "obsidian";
-import AIToolboxPlugin from "../main";
+import LLMToolboxPlugin from "../main";
 import {
 	AIProviderConfig,
 	AIProviderType,
@@ -156,7 +156,7 @@ async function testModel(provider: AIProviderConfig, model: AIModelConfig): Prom
  */
 export function displayProvidersSettings(
 	containerEl: HTMLElement,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	callbacks: ProviderSettingsCallbacks
 ): void {
 	// Add provider button
@@ -188,7 +188,7 @@ export function displayProvidersSettings(
 
 function displayProviderSettings(
 	containerEl: HTMLElement,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	provider: AIProviderConfig,
 	callbacks: ProviderSettingsCallbacks
 ): void {
@@ -306,7 +306,7 @@ function displayProviderSettings(
 
 function displayModelSettings(
 	containerEl: HTMLElement,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	provider: AIProviderConfig,
 	model: AIModelConfig,
 	callbacks: ProviderSettingsCallbacks
diff --git a/src/settings/types.ts b/src/settings/types.ts
index 77fcdc1..5c9c758 100644
--- a/src/settings/types.ts
+++ b/src/settings/types.ts
@@ -152,7 +152,7 @@ export const DEFAULT_WORKFLOW_CONFIG: Omit<WorkflowConfig, 'id'> = {
 	timestampGranularity: 'disabled'
 };
 
-export interface AIToolboxSettings {
+export interface LLMToolboxSettings {
 	impersonateBrowser: string;
 	ytdlpLocation: string;
 	ffmpegLocation: string;
@@ -171,7 +171,7 @@ export function generateId(): string {
 	return Math.random().toString(36).substring(2, 11);
 }
 
-export const DEFAULT_SETTINGS: AIToolboxSettings = {
+export const DEFAULT_SETTINGS: LLMToolboxSettings = {
 	impersonateBrowser: 'chrome',
 	ytdlpLocation: '',
 	ffmpegLocation: '',
diff --git a/src/settings/workflows.ts b/src/settings/workflows.ts
index 1903cd7..ffe82c7 100644
--- a/src/settings/workflows.ts
+++ b/src/settings/workflows.ts
@@ -1,5 +1,5 @@
 import { Notice, setIcon, Setting } from "obsidian";
-import AIToolboxPlugin from "../main";
+import LLMToolboxPlugin from "../main";
 import {
 	WorkflowConfig,
 	WorkflowOutputType,
@@ -56,7 +56,7 @@ const PROMPT_SOURCE_OPTIONS: Record<PromptSourceType, string> = {
  */
 export function displayWorkflowsSettings(
 	containerEl: HTMLElement,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	callbacks: WorkflowSettingsCallbacks
 ): void {
 	// Add workflow button
@@ -90,7 +90,7 @@ export function displayWorkflowsSettings(
 
 function displayWorkflowSettings(
 	containerEl: HTMLElement,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	workflow: WorkflowConfig,
 	callbacks: WorkflowSettingsCallbacks
 ): void {
@@ -154,7 +154,7 @@ function displayWorkflowSettings(
  */
 function displayChatWorkflowSettings(
 	contentContainer: HTMLElement,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	workflow: WorkflowConfig,
 	callbacks: WorkflowSettingsCallbacks,
 	isExpanded: () => boolean
@@ -295,7 +295,7 @@ function displayChatWorkflowSettings(
 
 function displayWorkflowProviderSelection(
 	containerEl: HTMLElement,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	workflow: WorkflowConfig
 ): void {
 	const providers = plugin.settings.providers;
@@ -344,7 +344,7 @@ function displayWorkflowProviderSelection(
  */
 function displayTranscriptionProviderSelection(
 	containerEl: HTMLElement,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	workflow: WorkflowConfig
 ): void {
 	const providers = plugin.settings.providers;
@@ -392,7 +392,7 @@ function displayTranscriptionProviderSelection(
  */
 function displayTranscriptionWorkflowSettings(
 	contentContainer: HTMLElement,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	workflow: WorkflowConfig,
 	callbacks: WorkflowSettingsCallbacks,
 	isExpanded: () => boolean
@@ -553,7 +553,7 @@ function displayTranscriptionWorkflowSettings(
  */
 function displayContextSection(
 	contentContainer: HTMLElement,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	workflow: WorkflowConfig,
 	callbacks: WorkflowSettingsCallbacks,
 	isExpanded: () => boolean
@@ -638,7 +638,7 @@ function displayContextSection(
  */
 function displayWorkflowContextSection(
 	containerEl: HTMLElement,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	workflow: WorkflowConfig,
 	callbacks: WorkflowSettingsCallbacks,
 	isExpanded: () => boolean
@@ -774,7 +774,7 @@ function renderTokenReference(
 function displayAvailableTokensSection(
 	containerEl: HTMLElement,
 	workflow: WorkflowConfig,
-	plugin: AIToolboxPlugin,
+	plugin: LLMToolboxPlugin,
 	callbacks: WorkflowSettingsCallbacks
 ): void {
 	const contexts = workflow.contexts ?? [];
diff --git a/src/tokens/types.ts b/src/tokens/types.ts
index f2c2947..a18b6a8 100644
--- a/src/tokens/types.ts
+++ b/src/tokens/types.ts
@@ -95,15 +95,16 @@ export interface TokenDefinitionOptions {
 /**
  * Get token definitions for a workflow type.
  * For transcription workflows, the transcriptionWithTimestamps token is excluded
- * when timestampGranularity is 'disabled'.
+ * when timestampGranularity is 'disabled' or undefined (default).
  */
 export function getTokenDefinitionsForType(
 	type: WorkflowType,
 	options?: TokenDefinitionOptions
 ): TokenDefinition[] {
 	if (type === 'transcription') {
-		// Filter out transcriptionWithTimestamps when granularity is disabled
-		if (options?.timestampGranularity === 'disabled') {
+		// Filter out transcriptionWithTimestamps when granularity is disabled or not set
+		const granularity = options?.timestampGranularity ?? 'disabled';
+		if (granularity === 'disabled') {
 			return TRANSCRIPTION_WORKFLOW_TOKENS.filter(
 				token => token.name !== 'transcriptionWithTimestamps'
 			);

From 569f97746372d977c7e35a8f7acf8f586af6a900 Mon Sep 17 00:00:00 2001
From: dalinicus <contact@dalinicus.com>
Date: Fri, 9 Jan 2026 21:04:42 -0600
Subject: [PATCH 3/7] eh changed my mind

---
 README.md                           | 12 ++++++------
 manifest.json                       |  8 ++++----
 src/handlers/input/types.ts         |  4 ++--
 src/main.ts                         | 10 +++++-----
 src/processing/video-processor.ts   | 10 +++++-----
 src/processing/workflow-chaining.ts |  6 +++---
 src/processing/workflow-executor.ts | 16 ++++++++--------
 src/providers/provider-factory.ts   |  6 +++---
 src/settings/additional-settings.ts |  4 ++--
 src/settings/index.ts               | 10 +++++-----
 src/settings/providers.ts           |  8 ++++----
 src/settings/types.ts               |  4 ++--
 src/settings/workflows.ts           | 20 ++++++++++----------
 13 files changed, 59 insertions(+), 59 deletions(-)

diff --git a/README.md b/README.md
index b3fe3ae..91b9b0d 100644
--- a/README.md
+++ b/README.md
@@ -1,6 +1,6 @@
-# LLM Toolbox
+# AI Toolbox
 
-A personal collection of LLM tools to enhance the Obsidian.md workflow.
+A personal collection of AI tools to enhance the Obsidian.md workflow.
 
 ## Features
 
@@ -37,7 +37,7 @@ pip install yt-dlp[curl_cffi]
 ## How to Use
 
 1. **Configure the plugin**:
-   - Go to **Settings → LLM Toolbox**
+   - Go to **Settings → AI Toolbox**
    - Enter your Azure OpenAI endpoint, API key, and deployment name
    - Configure optional settings (timestamps, language, output folder)
 
@@ -53,14 +53,14 @@ pip install yt-dlp[curl_cffi]
 
 1. Install the [BRAT plugin](https://github.com/TfTHacker/obsidian42-brat) if you haven't already
 2. Open **Settings → BRAT → Add Beta plugin**
-3. Enter this repository ID: `dalinicus/obsidian-llm-toolbox`
+3. Enter this repository ID: `dalinicus/obsidian-ai-toolbox`
 4. Click **Add Plugin**
 5. Enable the plugin in **Settings → Community plugins**
 
 ### Manual Installation
 
-1. Download the latest release from the [Releases page](https://github.com/dalinicus/obsidian-llm-toolbox/releases)
-2. Extract the files to your vault's plugins folder: `VaultFolder/.obsidian/plugins/llm-toolbox/`
+1. Download the latest release from the [Releases page](https://github.com/dalinicus/obsidian-ai-toolbox/releases)
+2. Extract the files to your vault's plugins folder: `VaultFolder/.obsidian/plugins/ai-toolbox/`
 3. Reload Obsidian
 4. Enable the plugin in **Settings → Community plugins**
 
diff --git a/manifest.json b/manifest.json
index ba922aa..de98265 100644
--- a/manifest.json
+++ b/manifest.json
@@ -1,9 +1,9 @@
 {
-	"id": "llm-toolbox",
-	"name": "LLM Toolbox",
-	"version": "1.0.5",
+	"id": "ai-toolbox",
+	"name": "AI Toolbox",
+	"version": "1.0.3",
 	"minAppVersion": "0.15.0",
-	"description": "Personal collection of LLM tools to enhance the Obsidian.md workflow.",
+	"description": "Personal collection of AI tools to enhance the Obsidian.md workflow.",
 	"author": "dalinicus",
 	"authorUrl": "https://github.com/dalinicus",
 	"fundingUrl": "https://buymeacoffee.com/dalinicus",
diff --git a/src/handlers/input/types.ts b/src/handlers/input/types.ts
index deb19a0..6e0d6e7 100644
--- a/src/handlers/input/types.ts
+++ b/src/handlers/input/types.ts
@@ -1,5 +1,5 @@
 import { App } from 'obsidian';
-import { LLMToolboxSettings, WorkflowConfig } from '../../settings';
+import { AIToolboxSettings, WorkflowConfig } from '../../settings';
 import { VideoMetadata } from '../../processing';
 
 /**
@@ -22,7 +22,7 @@ export interface InputContext {
     /** The Obsidian App instance */
     app: App;
     /** Plugin settings */
-    settings: LLMToolboxSettings;
+    settings: AIToolboxSettings;
     /** The workflow configuration being executed */
     workflow: WorkflowConfig;
 }
diff --git a/src/main.ts b/src/main.ts
index c32deaf..4e3b00c 100644
--- a/src/main.ts
+++ b/src/main.ts
@@ -1,10 +1,10 @@
 import { Notice, Plugin } from 'obsidian';
-import { DEFAULT_SETTINGS, LLMToolboxSettings, LLMToolboxSettingTab } from "./settings/index";
+import { DEFAULT_SETTINGS, AIToolboxSettings, AIToolboxSettingTab } from "./settings/index";
 import { WorkflowSuggesterModal } from "./components/workflow-suggester";
 import { executeWorkflow } from "./processing/workflow-executor";
 
-export default class LLMToolboxPlugin extends Plugin {
-	settings: LLMToolboxSettings;
+export default class AIToolboxPlugin extends Plugin {
+	settings: AIToolboxSettings;
 
 	async onload() {
 		await this.loadSettings();
@@ -17,7 +17,7 @@ export default class LLMToolboxPlugin extends Plugin {
 		});
 
 		// This adds a settings tab so the user can configure various aspects of the plugin
-		this.addSettingTab(new LLMToolboxSettingTab(this.app, this));
+		this.addSettingTab(new AIToolboxSettingTab(this.app, this));
 	}
 
 	/**
@@ -38,7 +38,7 @@ export default class LLMToolboxPlugin extends Plugin {
 	}
 
 	async loadSettings() {
-		this.settings = Object.assign({}, DEFAULT_SETTINGS, await this.loadData() as Partial<LLMToolboxSettings>);
+		this.settings = Object.assign({}, DEFAULT_SETTINGS, await this.loadData() as Partial<AIToolboxSettings>);
 	}
 
 	async saveSettings() {
diff --git a/src/processing/video-processor.ts b/src/processing/video-processor.ts
index 585c3eb..7e3e6f7 100644
--- a/src/processing/video-processor.ts
+++ b/src/processing/video-processor.ts
@@ -4,7 +4,7 @@ import * as os from 'os';
 import * as fs from 'fs';
 import { Buffer } from 'buffer';
 import { Notice } from 'obsidian';
-import { LLMToolboxSettings } from '../settings';
+import { AIToolboxSettings } from '../settings';
 import { videoPlatformRegistry, VideoMetadata } from './video-platforms';
 
 /**
@@ -198,9 +198,9 @@ export interface ExtractAudioResult {
 }
 
 /**
- * Converts LLMToolboxSettings to VideoProcessorConfig.
+ * Converts AIToolboxSettings to VideoProcessorConfig.
  */
-function settingsToProcessorConfig(settings: LLMToolboxSettings): VideoProcessorConfig {
+function settingsToProcessorConfig(settings: AIToolboxSettings): VideoProcessorConfig {
     return {
         ytdlpLocation: settings.ytdlpLocation,
         ffmpegLocation: settings.ffmpegLocation,
@@ -216,7 +216,7 @@ function settingsToProcessorConfig(settings: LLMToolboxSettings): VideoProcessor
  * Uses yt-dlp directly via child_process for audio extraction.
  * Requires yt-dlp and ffmpeg to be installed and available in PATH.
  */
-export async function extractAudioFromUrl(url: string, settings: LLMToolboxSettings): Promise<ExtractAudioResult | null> {
+export async function extractAudioFromUrl(url: string, settings: AIToolboxSettings): Promise<ExtractAudioResult | null> {
     try {
         if (!url || !url.trim()) {
             new Notice('URL is empty');
@@ -270,7 +270,7 @@ export async function extractAudioFromUrl(url: string, settings: LLMToolboxSetti
  * Uses yt-dlp directly via child_process for audio extraction.
  * Requires yt-dlp and ffmpeg to be installed and available in PATH.
  */
-export async function extractAudioFromClipboard(settings: LLMToolboxSettings): Promise<ExtractAudioResult | null> {
+export async function extractAudioFromClipboard(settings: AIToolboxSettings): Promise<ExtractAudioResult | null> {
     try {
         const clipboardText = await navigator.clipboard.readText();
 
diff --git a/src/processing/workflow-chaining.ts b/src/processing/workflow-chaining.ts
index 3d9189d..c660f9b 100644
--- a/src/processing/workflow-chaining.ts
+++ b/src/processing/workflow-chaining.ts
@@ -1,5 +1,5 @@
 import { App, MarkdownView } from 'obsidian';
-import { WorkflowConfig, LLMToolboxSettings, TimestampGranularity } from '../settings';
+import { WorkflowConfig, AIToolboxSettings, TimestampGranularity } from '../settings';
 import { TranscriptionResult, TranscriptionChunk } from '../providers';
 
 /**
@@ -30,7 +30,7 @@ export interface DependencyExecutionContext {
     /** The Obsidian App instance */
     app: App;
     /** Plugin settings */
-    settings: LLMToolboxSettings;
+    settings: AIToolboxSettings;
     /** Results from executed dependencies */
     results: WorkflowResultsMap;
     /** Workflow IDs currently in the execution stack (for circular detection) */
@@ -43,7 +43,7 @@ export interface DependencyExecutionContext {
  */
 export function detectCircularDependency(
     workflow: WorkflowConfig,
-    settings: LLMToolboxSettings,
+    settings: AIToolboxSettings,
     visited: Set<string> = new Set(),
     path: string[] = []
 ): string[] {
diff --git a/src/processing/workflow-executor.ts b/src/processing/workflow-executor.ts
index e11e175..af5c6f8 100644
--- a/src/processing/workflow-executor.ts
+++ b/src/processing/workflow-executor.ts
@@ -1,5 +1,5 @@
 import { App, Notice, TFile } from 'obsidian';
-import { WorkflowConfig, LLMToolboxSettings, TranscriptionSourceType } from '../settings';
+import { WorkflowConfig, AIToolboxSettings, TranscriptionSourceType } from '../settings';
 import { createWorkflowProvider, ChatMessage, TranscriptionOptions, TranscriptionResult } from '../providers';
 import { videoPlatformRegistry } from './video-platforms';
 import { generateFilenameTimestamp } from '../utils/date-utils';
@@ -130,7 +130,7 @@ async function getPromptText(app: App, workflow: WorkflowConfig): Promise<string
  */
 async function executeDependencies(
 	app: App,
-	settings: LLMToolboxSettings,
+	settings: AIToolboxSettings,
 	workflow: WorkflowConfig,
 	results: WorkflowResultsMap,
 	executionStack: Set<string>
@@ -194,7 +194,7 @@ async function executeDependencies(
  */
 export async function executeWorkflow(
 	app: App,
-	settings: LLMToolboxSettings,
+	settings: AIToolboxSettings,
 	workflow: WorkflowConfig
 ): Promise<void> {
 	// Check for circular dependencies before starting
@@ -242,7 +242,7 @@ export async function executeWorkflow(
  */
 async function executeWorkflowInternal(
 	app: App,
-	settings: LLMToolboxSettings,
+	settings: AIToolboxSettings,
 	workflow: WorkflowConfig,
 	dependencyResults: WorkflowResultsMap
 ): Promise<WorkflowExecutionResult> {
@@ -260,7 +260,7 @@ async function executeWorkflowInternal(
  */
 async function executeChatWorkflow(
 	app: App,
-	settings: LLMToolboxSettings,
+	settings: AIToolboxSettings,
 	workflow: WorkflowConfig,
 	dependencyResults?: WorkflowResultsMap
 ): Promise<void> {
@@ -337,7 +337,7 @@ async function executeChatWorkflow(
  */
 async function executeChatWorkflowInternal(
 	app: App,
-	settings: LLMToolboxSettings,
+	settings: AIToolboxSettings,
 	workflow: WorkflowConfig,
 	dependencyResults: WorkflowResultsMap
 ): Promise<WorkflowExecutionResult> {
@@ -405,7 +405,7 @@ async function executeChatWorkflowInternal(
  */
 async function executeTranscriptionWorkflow(
 	app: App,
-	settings: LLMToolboxSettings,
+	settings: AIToolboxSettings,
 	workflow: WorkflowConfig,
 	_dependencyResults?: WorkflowResultsMap
 ): Promise<void> {
@@ -467,7 +467,7 @@ async function executeTranscriptionWorkflow(
  */
 async function executeTranscriptionWorkflowInternal(
 	app: App,
-	settings: LLMToolboxSettings,
+	settings: AIToolboxSettings,
 	workflow: WorkflowConfig,
 	_dependencyResults: WorkflowResultsMap
 ): Promise<WorkflowExecutionResult> {
diff --git a/src/providers/provider-factory.ts b/src/providers/provider-factory.ts
index e0c0277..cbef66b 100644
--- a/src/providers/provider-factory.ts
+++ b/src/providers/provider-factory.ts
@@ -1,4 +1,4 @@
-import { LLMToolboxSettings, AIProviderConfig, AIModelConfig, DEFAULT_OPENAI_ENDPOINT, WorkflowConfig, ProviderModelSelection } from '../settings/index';
+import { AIToolboxSettings, AIProviderConfig, AIModelConfig, DEFAULT_OPENAI_ENDPOINT, WorkflowConfig, ProviderModelSelection } from '../settings/index';
 import { ModelProvider, ModelProviderConfig } from './types';
 import { AzureOpenAIModelProvider } from './azure-openai-provider';
 import { OpenAIModelProvider } from './openai-provider';
@@ -62,7 +62,7 @@ export function createModelProvider(config: ModelProviderConfig): ModelProvider
  * @returns A configured ModelProvider, or null if not found
  * @throws ProviderCreationError if the provider cannot be created
  */
-function createProviderFromSelection(settings: LLMToolboxSettings, selection: ProviderModelSelection): ModelProvider | null {
+function createProviderFromSelection(settings: AIToolboxSettings, selection: ProviderModelSelection): ModelProvider | null {
 	const provider = settings.providers.find(p => p.id === selection.providerId);
 	if (!provider) {
 		return null;
@@ -85,7 +85,7 @@ function createProviderFromSelection(settings: LLMToolboxSettings, selection: Pr
  * @returns A configured ModelProvider for the workflow, or null if not configured
  * @throws ProviderCreationError if the provider cannot be created
  */
-export function createWorkflowProvider(settings: LLMToolboxSettings, workflow: WorkflowConfig): ModelProvider | null {
+export function createWorkflowProvider(settings: AIToolboxSettings, workflow: WorkflowConfig): ModelProvider | null {
 	if (!workflow.provider) {
 		return null;
 	}
diff --git a/src/settings/additional-settings.ts b/src/settings/additional-settings.ts
index cc57536..9e413f8 100644
--- a/src/settings/additional-settings.ts
+++ b/src/settings/additional-settings.ts
@@ -1,5 +1,5 @@
 import { Setting } from "obsidian";
-import LLMToolboxPlugin from "../main";
+import AIToolboxPlugin from "../main";
 
 /**
  * Callbacks for the additional settings tab to communicate with the main settings tab
@@ -14,7 +14,7 @@ export interface AdditionalSettingsCallbacks {
  */
 export function displayAdditionalSettings(
 	containerEl: HTMLElement,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	callbacks: AdditionalSettingsCallbacks
 ): void {
 	const showAdvanced = callbacks.isAdvancedVisible();
diff --git a/src/settings/index.ts b/src/settings/index.ts
index 6ec9093..77e48a2 100644
--- a/src/settings/index.ts
+++ b/src/settings/index.ts
@@ -1,5 +1,5 @@
 import { App, PluginSettingTab, Setting } from "obsidian";
-import LLMToolboxPlugin from "../main";
+import AIToolboxPlugin from "../main";
 import { SettingsTabType, ExpandOnNextRenderState } from "./types";
 import { displayProvidersSettings, ProviderSettingsCallbacks } from "./providers";
 import { displayWorkflowsSettings, WorkflowSettingsCallbacks } from "./workflows";
@@ -22,20 +22,20 @@ export type {
 	TimestampGranularity,
 	ChatContextType,
 	ChatContextConfig,
-	LLMToolboxSettings,
+	AIToolboxSettings,
 	SettingsTabType,
 	ExpandOnNextRenderState
 } from "./types";
 
-export class LLMToolboxSettingTab extends PluginSettingTab {
-	plugin: LLMToolboxPlugin;
+export class AIToolboxSettingTab extends PluginSettingTab {
+	plugin: AIToolboxPlugin;
 	private activeTab: SettingsTabType = 'providers';
 	// Track IDs that should start expanded on next render (cleared after use)
 	private expandOnNextRender: ExpandOnNextRenderState = {};
 	// Global advanced settings visibility state
 	private showAdvancedSettings = false;
 
-	constructor(app: App, plugin: LLMToolboxPlugin) {
+	constructor(app: App, plugin: AIToolboxPlugin) {
 		super(app, plugin);
 		this.plugin = plugin;
 	}
diff --git a/src/settings/providers.ts b/src/settings/providers.ts
index 30c9b3c..76b2233 100644
--- a/src/settings/providers.ts
+++ b/src/settings/providers.ts
@@ -1,5 +1,5 @@
 import { Setting, setIcon, ButtonComponent, Notice } from "obsidian";
-import LLMToolboxPlugin from "../main";
+import AIToolboxPlugin from "../main";
 import {
 	AIProviderConfig,
 	AIProviderType,
@@ -156,7 +156,7 @@ async function testModel(provider: AIProviderConfig, model: AIModelConfig): Prom
  */
 export function displayProvidersSettings(
 	containerEl: HTMLElement,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	callbacks: ProviderSettingsCallbacks
 ): void {
 	// Add provider button
@@ -188,7 +188,7 @@ export function displayProvidersSettings(
 
 function displayProviderSettings(
 	containerEl: HTMLElement,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	provider: AIProviderConfig,
 	callbacks: ProviderSettingsCallbacks
 ): void {
@@ -306,7 +306,7 @@ function displayProviderSettings(
 
 function displayModelSettings(
 	containerEl: HTMLElement,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	provider: AIProviderConfig,
 	model: AIModelConfig,
 	callbacks: ProviderSettingsCallbacks
diff --git a/src/settings/types.ts b/src/settings/types.ts
index 5c9c758..77fcdc1 100644
--- a/src/settings/types.ts
+++ b/src/settings/types.ts
@@ -152,7 +152,7 @@ export const DEFAULT_WORKFLOW_CONFIG: Omit<WorkflowConfig, 'id'> = {
 	timestampGranularity: 'disabled'
 };
 
-export interface LLMToolboxSettings {
+export interface AIToolboxSettings {
 	impersonateBrowser: string;
 	ytdlpLocation: string;
 	ffmpegLocation: string;
@@ -171,7 +171,7 @@ export function generateId(): string {
 	return Math.random().toString(36).substring(2, 11);
 }
 
-export const DEFAULT_SETTINGS: LLMToolboxSettings = {
+export const DEFAULT_SETTINGS: AIToolboxSettings = {
 	impersonateBrowser: 'chrome',
 	ytdlpLocation: '',
 	ffmpegLocation: '',
diff --git a/src/settings/workflows.ts b/src/settings/workflows.ts
index ffe82c7..1903cd7 100644
--- a/src/settings/workflows.ts
+++ b/src/settings/workflows.ts
@@ -1,5 +1,5 @@
 import { Notice, setIcon, Setting } from "obsidian";
-import LLMToolboxPlugin from "../main";
+import AIToolboxPlugin from "../main";
 import {
 	WorkflowConfig,
 	WorkflowOutputType,
@@ -56,7 +56,7 @@ const PROMPT_SOURCE_OPTIONS: Record<PromptSourceType, string> = {
  */
 export function displayWorkflowsSettings(
 	containerEl: HTMLElement,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	callbacks: WorkflowSettingsCallbacks
 ): void {
 	// Add workflow button
@@ -90,7 +90,7 @@ export function displayWorkflowsSettings(
 
 function displayWorkflowSettings(
 	containerEl: HTMLElement,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	workflow: WorkflowConfig,
 	callbacks: WorkflowSettingsCallbacks
 ): void {
@@ -154,7 +154,7 @@ function displayWorkflowSettings(
  */
 function displayChatWorkflowSettings(
 	contentContainer: HTMLElement,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	workflow: WorkflowConfig,
 	callbacks: WorkflowSettingsCallbacks,
 	isExpanded: () => boolean
@@ -295,7 +295,7 @@ function displayChatWorkflowSettings(
 
 function displayWorkflowProviderSelection(
 	containerEl: HTMLElement,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	workflow: WorkflowConfig
 ): void {
 	const providers = plugin.settings.providers;
@@ -344,7 +344,7 @@ function displayWorkflowProviderSelection(
  */
 function displayTranscriptionProviderSelection(
 	containerEl: HTMLElement,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	workflow: WorkflowConfig
 ): void {
 	const providers = plugin.settings.providers;
@@ -392,7 +392,7 @@ function displayTranscriptionProviderSelection(
  */
 function displayTranscriptionWorkflowSettings(
 	contentContainer: HTMLElement,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	workflow: WorkflowConfig,
 	callbacks: WorkflowSettingsCallbacks,
 	isExpanded: () => boolean
@@ -553,7 +553,7 @@ function displayTranscriptionWorkflowSettings(
  */
 function displayContextSection(
 	contentContainer: HTMLElement,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	workflow: WorkflowConfig,
 	callbacks: WorkflowSettingsCallbacks,
 	isExpanded: () => boolean
@@ -638,7 +638,7 @@ function displayContextSection(
  */
 function displayWorkflowContextSection(
 	containerEl: HTMLElement,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	workflow: WorkflowConfig,
 	callbacks: WorkflowSettingsCallbacks,
 	isExpanded: () => boolean
@@ -774,7 +774,7 @@ function renderTokenReference(
 function displayAvailableTokensSection(
 	containerEl: HTMLElement,
 	workflow: WorkflowConfig,
-	plugin: LLMToolboxPlugin,
+	plugin: AIToolboxPlugin,
 	callbacks: WorkflowSettingsCallbacks
 ): void {
 	const contexts = workflow.contexts ?? [];

From 511ed78690cfc9c778625702057e46069c2f6202 Mon Sep 17 00:00:00 2001
From: dalinicus <contact@dalinicus.com>
Date: Sun, 11 Jan 2026 19:04:22 -0600
Subject: [PATCH 4/7] clean up picker

---
 AGENTS.md                     |  31 +++
 src/components/path-picker.ts | 373 +++++++++++-----------------------
 src/settings/types.ts         |   2 -
 src/settings/workflows.ts     |  36 ++--
 4 files changed, 162 insertions(+), 280 deletions(-)

diff --git a/AGENTS.md b/AGENTS.md
index b72af5f..6212a38 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -123,6 +123,37 @@ npm run build
 - Attach `manifest.json`, `main.js`, and `styles.css` (if present) to the release as individual assets.
 - After the initial release, follow the process to add/update your plugin in the community catalog as required.
 
+### Release preparation checklist
+
+Before creating a GitHub release, follow these steps to update version files:
+
+#### 1. Update manifest.json
+- Bump the `version` field in `manifest.json` following Semantic Versioning (x.y.z format)
+- Ensure the `minAppVersion` is accurate for any new Obsidian APIs used
+
+#### 2. Update versions.json
+The `versions.json` file maps plugin versions to minimum Obsidian app versions. Update it based on the release type:
+
+**For patch releases (x.y.Z):**
+- Add the new version entry: `"x.y.z": "minimum-app-version"`
+- Keep all existing entries unchanged
+
+**For minor releases (x.Y.0):**
+- Add the new version entry: `"x.y.0": "minimum-app-version"`
+- For the previous minor version (x.Y-1.*), keep only the latest patch version
+- Remove all other patch versions for that minor version
+- Example: If releasing 1.2.0, keep only 1.1.3 (remove 1.1.0, 1.1.1, 1.1.2)
+
+**For major releases (X.0.0):**
+- Add the new version entry: `"x.0.0": "minimum-app-version"`
+- For the previous major version (X-1.*.*), keep the latest two minor versions, each with their latest patch version
+- Remove all other minor and patch versions for that major version
+- Example: If releasing 2.0.0, keep only 1.2.3 and 1.3.2 (remove 1.0.0, 1.1.0, 1.2.1, 1.2.2, 1.3.0, 1.3.1, etc.)
+
+#### 3. Run linting
+- Run `eslint ./src/` to check for code quality issues
+- Fix any errors or warnings before creating the release
+
 ## Security, privacy, and compliance
 
 Follow Obsidian's **Developer Policies** and **Plugin Guidelines**. In particular:
diff --git a/src/components/path-picker.ts b/src/components/path-picker.ts
index 715adc4..d63bc9e 100644
--- a/src/components/path-picker.ts
+++ b/src/components/path-picker.ts
@@ -1,328 +1,189 @@
-import { App, Setting, AbstractInputSuggest, TFolder, TFile } from "obsidian";
+import { App, Setting, AbstractInputSuggest, TFolder, TFile, TAbstractFile } from "obsidian";
 
 /**
- * Internal folder suggestion component.
+ * Type of path selection.
  */
-class FolderSuggestInternal extends AbstractInputSuggest<TFolder> {
-    private textInputEl: HTMLInputElement;
-
-    constructor(app: App, inputEl: HTMLInputElement) {
-        super(app, inputEl);
-        this.textInputEl = inputEl;
-    }
-
-    getSuggestions(inputStr: string): TFolder[] {
-        const inputLower = inputStr.toLowerCase().trim();
-        const allFolders = this.getAllFolders();
-
-        if (inputLower === "") {
-            return allFolders;
-        }
-
-        return allFolders.filter(folder =>
-            folder.path.toLowerCase().includes(inputLower)
-        );
-    }
+export type PathSelectionType = "folder" | "file";
 
-    renderSuggestion(folder: TFolder, el: HTMLElement): void {
-        const displayPath = folder.path === "" ? "/" : folder.path;
-        el.createEl("div", { text: displayPath, cls: "folder-suggest-item" });
-    }
-
-    selectSuggestion(folder: TFolder): void {
-        this.textInputEl.value = folder.path;
-        this.textInputEl.dispatchEvent(new Event("input", { bubbles: true }));
-        this.close();
-    }
-
-    private getAllFolders(): TFolder[] {
-        const folders: TFolder[] = [];
-        
-        const rootFolder = this.app.vault.getRoot();
-        if (rootFolder) {
-            folders.push(rootFolder);
-        }
-
-        const allFolders = this.app.vault.getAllFolders();
-        folders.push(...allFolders);
-
-        folders.sort((a, b) => {
-            if (a.path === "") return -1;
-            if (b.path === "") return 1;
-            return a.path.localeCompare(b.path);
-        });
-
-        return folders;
-    }
+/**
+ * Represents a path item (folder or file) in the unified picker.
+ */
+interface PathItem {
+    type: PathSelectionType;
+    path: string;
+    displayPath: string;
+    item: TAbstractFile;
 }
 
 /**
- * Internal file suggestion component that filters by folder.
+ * Unified path suggestion component that shows both folders and files.
  */
-class FileSuggestInternal extends AbstractInputSuggest<TFile> {
+class UnifiedPathSuggest extends AbstractInputSuggest<PathItem> {
     private textInputEl: HTMLInputElement;
-    private folderPath: string = "";
-
-    constructor(app: App, inputEl: HTMLInputElement) {
+    private includeFiles: boolean;
+    private onPathSelected: (path: string, type: PathSelectionType) => void;
+
+    constructor(
+        app: App,
+        inputEl: HTMLInputElement,
+        includeFiles: boolean,
+        onPathSelected: (path: string, type: PathSelectionType) => void
+    ) {
         super(app, inputEl);
         this.textInputEl = inputEl;
+        this.includeFiles = includeFiles;
+        this.onPathSelected = onPathSelected;
     }
 
-    setFolderPath(folderPath: string): void {
-        this.folderPath = folderPath;
-    }
-
-    getSuggestions(inputStr: string): TFile[] {
+    getSuggestions(inputStr: string): PathItem[] {
         const inputLower = inputStr.toLowerCase().trim();
-        const files = this.getFilesInFolder();
+        const items = this.getAllItems();
 
         if (inputLower === "") {
-            return files;
+            return items;
         }
 
-        return files.filter(file => {
-            const displayPath = this.getDisplayPath(file);
-            return displayPath.toLowerCase().includes(inputLower);
-        });
+        return items.filter(item =>
+            item.path.toLowerCase().includes(inputLower)
+        );
     }
 
-    renderSuggestion(file: TFile, el: HTMLElement): void {
-        const displayPath = this.getDisplayPath(file);
-        el.createEl("div", { text: displayPath, cls: "file-suggest-item" });
+    renderSuggestion(item: PathItem, el: HTMLElement): void {
+        el.addClass("path-picker-suggestion");
+        el.createEl("span", {
+            text: item.displayPath,
+            cls: `path-picker-${item.type}`
+        });
     }
 
-    selectSuggestion(file: TFile): void {
-        this.textInputEl.value = file.name;
-        this.textInputEl.dataset.fullPath = file.path;
-        this.textInputEl.dispatchEvent(new CustomEvent("file-selected", {
-            bubbles: true,
-            detail: { path: file.path, name: file.name }
-        }));
+    selectSuggestion(item: PathItem): void {
+        this.textInputEl.value = item.path;
+        this.textInputEl.dataset.selectionType = item.type;
+        this.onPathSelected(item.path, item.type);
         this.close();
     }
 
-    private getDisplayPath(file: TFile): string {
-        if (this.folderPath && file.path.startsWith(this.folderPath + "/")) {
-            return file.path.substring(this.folderPath.length + 1);
+    private getAllItems(): PathItem[] {
+        const items: PathItem[] = [];
+
+        // Add folders
+        const rootFolder = this.app.vault.getRoot();
+        if (rootFolder) {
+            items.push({
+                type: "folder",
+                path: "",
+                displayPath: "/ (root)",
+                item: rootFolder
+            });
         }
-        return file.path;
-    }
 
-    private getFilesInFolder(): TFile[] {
-        const allFiles = this.app.vault.getFiles();
-        
-        if (!this.folderPath) {
-            return allFiles.sort((a, b) => a.path.localeCompare(b.path));
+        for (const folder of this.app.vault.getAllFolders()) {
+            items.push({
+                type: "folder",
+                path: folder.path,
+                displayPath: folder.path,
+                item: folder
+            });
         }
 
-        const normalizedFolder = this.folderPath.replace(/\/$/, "");
-        const folder = this.app.vault.getAbstractFileByPath(normalizedFolder);
-        
-        if (!(folder instanceof TFolder)) {
-            return [];
+        // Add markdown files if enabled
+        if (this.includeFiles) {
+            for (const file of this.app.vault.getMarkdownFiles()) {
+                // Display path without .md extension
+                const displayPath = file.path.replace(/\.md$/, "");
+                items.push({
+                    type: "file",
+                    path: file.path,
+                    displayPath: displayPath,
+                    item: file
+                });
+            }
         }
 
-        const filesInFolder = allFiles.filter(file => {
-            if (normalizedFolder === "") {
-                return true;
+        // Sort: folders first (root at top), then files, alphabetically within each group
+        items.sort((a, b) => {
+            if (a.type !== b.type) {
+                return a.type === "folder" ? -1 : 1;
             }
-            return file.path.startsWith(normalizedFolder + "/");
+            // Root folder always first
+            if (a.path === "") return -1;
+            if (b.path === "") return 1;
+            return a.path.localeCompare(b.path);
         });
 
-        filesInFolder.sort((a, b) => a.path.localeCompare(b.path));
-        return filesInFolder;
+        return items;
     }
 }
 
 /**
- * Mode for the path picker component.
- */
-export type PathPickerMode = "folder-only" | "folder-file";
-
-/**
- * Base options for the path picker.
+ * Options for the unified path picker.
  */
-interface PathPickerBaseOptions {
+export interface PathPickerOptions {
     containerEl: HTMLElement;
     app: App;
     name: string;
     description: string;
-    folderPlaceholder?: string;
-    initialFolderPath?: string;
-    onFolderChange?: (folderPath: string) => void;
+    placeholder?: string;
+    initialPath?: string;
+    /** If true, shows both folders and files; if false, only folders */
+    allowFiles?: boolean;
+    /** Called when the user selects a path */
+    onChange?: (path: string, type: PathSelectionType) => void;
 }
 
 /**
- * Options for folder-only mode.
+ * Result from creating a path picker.
  */
-export interface FolderOnlyPickerOptions extends PathPickerBaseOptions {
-    mode: "folder-only";
-}
-
-/**
- * Options for folder-file cascading mode.
- */
-export interface FolderFilePickerOptions extends PathPickerBaseOptions {
-    mode: "folder-file";
-    filePlaceholder?: string;
-    initialFilePath?: string;
-    onFileChange?: (filePath: string) => void;
-}
-
-export type PathPickerOptions = FolderOnlyPickerOptions | FolderFilePickerOptions;
-
-/**
- * Result for folder-only mode.
- */
-export interface FolderOnlyPickerResult {
-    setting: Setting;
-    folderInputEl: HTMLInputElement;
-}
-
-/**
- * Result for folder-file mode.
- */
-export interface FolderFilePickerResult {
+export interface PathPickerResult {
     setting: Setting;
-    folderInputEl: HTMLInputElement;
-    fileInputEl: HTMLInputElement;
-    fileSuggest: FileSuggestInternal;
+    inputEl: HTMLInputElement;
 }
 
-export type PathPickerResult<T extends PathPickerOptions> =
-    T extends FolderOnlyPickerOptions ? FolderOnlyPickerResult : FolderFilePickerResult;
-
 /**
- * Create a unified path picker component.
+ * Create a unified path picker component with a single search input.
  *
- * Supports two modes:
- * - "folder-only": Shows only a folder selection input
- * - "folder-file": Shows folder and file selection inputs side-by-side
+ * Shows folders and optionally files in one searchable list.
+ * Automatically determines selection type based on what the user picks.
  */
-export function createPathPicker<T extends PathPickerOptions>(options: T): PathPickerResult<T> {
-    if (options.mode === "folder-only") {
-        return createFolderOnlyPicker(options) as PathPickerResult<T>;
-    } else {
-        return createFolderFilePicker(options) as PathPickerResult<T>;
-    }
-}
-
-function createFolderOnlyPicker(options: FolderOnlyPickerOptions): FolderOnlyPickerResult {
+export function createPathPicker(options: PathPickerOptions): PathPickerResult {
     const {
         containerEl,
         app,
         name,
         description,
-        folderPlaceholder = "Select folder...",
-        initialFolderPath = "",
-        onFolderChange
+        placeholder = "Search...",
+        initialPath = "",
+        allowFiles = false,
+        onChange
     } = options;
 
-    let folderInputEl: HTMLInputElement;
+    let inputEl: HTMLInputElement;
 
     const setting = new Setting(containerEl)
         .setName(name)
         .setDesc(description)
-        .addSearch(folderSearch => {
-            folderInputEl = folderSearch.inputEl;
-            folderSearch
-                .setPlaceholder(folderPlaceholder)
-                .setValue(initialFolderPath)
-                .onChange((value) => {
-                    onFolderChange?.(value);
-                });
-
-            new FolderSuggestInternal(app, folderSearch.inputEl);
-            folderSearch.inputEl.addClass("path-picker-folder-input");
+        .addSearch(search => {
+            inputEl = search.inputEl;
+            search
+                .setPlaceholder(placeholder)
+                .setValue(initialPath);
+
+            new UnifiedPathSuggest(
+                app,
+                search.inputEl,
+                allowFiles,
+                (path, type) => {
+                    onChange?.(path, type);
+                }
+            );
+
+            search.inputEl.addClass("path-picker-input");
         });
 
     setting.settingEl.addClass("path-picker");
 
     return {
         setting,
-        folderInputEl: folderInputEl!
-    };
-}
-
-function createFolderFilePicker(options: FolderFilePickerOptions): FolderFilePickerResult {
-    const {
-        containerEl,
-        app,
-        name,
-        description,
-        folderPlaceholder = "Select folder...",
-        filePlaceholder = "Select file...",
-        initialFolderPath = "",
-        initialFilePath = "",
-        onFolderChange,
-        onFileChange
-    } = options;
-
-    let folderInputEl: HTMLInputElement;
-    let fileInputEl: HTMLInputElement;
-    let fileSuggest: FileSuggestInternal;
-
-    const setting = new Setting(containerEl)
-        .setName(name)
-        .setDesc(description)
-        .addSearch(folderSearch => {
-            folderInputEl = folderSearch.inputEl;
-            folderSearch
-                .setPlaceholder(folderPlaceholder)
-                .setValue(initialFolderPath)
-                .onChange((value) => {
-                    if (fileSuggest) {
-                        fileSuggest.setFolderPath(value);
-                    }
-                    // Enable/disable file input based on folder selection
-                    if (fileInputEl) {
-                        fileInputEl.disabled = !value;
-                        // Clear file if folder is cleared
-                        if (!value) {
-                            fileInputEl.value = "";
-                            fileInputEl.dataset.fullPath = "";
-                            onFileChange?.("");
-                        }
-                    }
-                    onFolderChange?.(value);
-                });
-
-            new FolderSuggestInternal(app, folderSearch.inputEl);
-            folderSearch.inputEl.addClass("path-picker-folder-input");
-        })
-        .addSearch(fileSearch => {
-            fileInputEl = fileSearch.inputEl;
-
-            // Extract just the filename from the initial path for display
-            const initialFileName = initialFilePath ? initialFilePath.split("/").pop() ?? "" : "";
-
-            fileSearch
-                .setPlaceholder(filePlaceholder)
-                .setValue(initialFileName);
-
-            // Disable file input if no folder is selected
-            fileSearch.inputEl.disabled = !initialFolderPath;
-
-            // Store the full path in dataset
-            fileSearch.inputEl.dataset.fullPath = initialFilePath;
-
-            // Listen for file selection (custom event with full path)
-            fileSearch.inputEl.addEventListener("file-selected", ((e: CustomEvent<{ path: string }>) => {
-                onFileChange?.(e.detail.path);
-            }) as EventListener);
-
-            fileSuggest = new FileSuggestInternal(app, fileSearch.inputEl);
-            fileSuggest.setFolderPath(initialFolderPath);
-            fileSearch.inputEl.addClass("path-picker-file-input");
-        });
-
-    setting.settingEl.addClass("path-picker", "path-picker-folder-file");
-
-    return {
-        setting,
-        folderInputEl: folderInputEl!,
-        fileInputEl: fileInputEl!,
-        fileSuggest: fileSuggest!
+        inputEl: inputEl!
     };
 }
-
diff --git a/src/settings/types.ts b/src/settings/types.ts
index 77fcdc1..51670bd 100644
--- a/src/settings/types.ts
+++ b/src/settings/types.ts
@@ -114,7 +114,6 @@ export interface WorkflowConfig {
 	type: WorkflowType;
 	promptText: string;
 	promptSourceType: PromptSourceType;
-	promptFolderPath: string;
 	promptFilePath: string;
 	provider: ProviderModelSelection | null;
 	outputType: WorkflowOutputType;
@@ -139,7 +138,6 @@ export const DEFAULT_WORKFLOW_CONFIG: Omit<WorkflowConfig, 'id'> = {
 	type: 'chat',
 	promptText: '',
 	promptSourceType: 'inline',
-	promptFolderPath: '',
 	promptFilePath: '',
 	provider: null,
 	outputType: 'popup',
diff --git a/src/settings/workflows.ts b/src/settings/workflows.ts
index 1903cd7..6fc8b94 100644
--- a/src/settings/workflows.ts
+++ b/src/settings/workflows.ts
@@ -204,21 +204,15 @@ function displayChatWorkflowSettings(
 	// Prompt file picker (only show when source is from-file)
 	if (promptSourceType === 'from-file') {
 		createPathPicker({
-			mode: 'folder-file',
 			containerEl: contentContainer,
 			app: plugin.app,
 			name: 'Prompt file',
-			description: 'First select a folder to filter, then select a file from that folder',
-			initialFolderPath: workflow.promptFolderPath ?? '',
-			initialFilePath: workflow.promptFilePath ?? '',
-			folderPlaceholder: 'Select folder...',
-			filePlaceholder: 'Select file...',
-			onFolderChange: async (folderPath: string) => {
-				workflow.promptFolderPath = folderPath;
-				await plugin.saveSettings();
-			},
-			onFileChange: async (filePath: string) => {
-				workflow.promptFilePath = filePath;
+			description: 'Search for a file to use as the prompt template',
+			placeholder: 'Search for file...',
+			initialPath: workflow.promptFilePath ?? '',
+			allowFiles: true,
+			onChange: async (path: string) => {
+				workflow.promptFilePath = path;
 				await plugin.saveSettings();
 			}
 		});
@@ -277,15 +271,14 @@ function displayChatWorkflowSettings(
 		// Output folder (only show if output type is new-note)
 		if (workflow.outputType === 'new-note') {
 			createPathPicker({
-				mode: 'folder-only',
 				containerEl: contentContainer,
 				app: plugin.app,
 				name: 'Output folder',
 				description: 'Folder where notes will be created (leave empty to use default)',
-				folderPlaceholder: 'Default folder',
-				initialFolderPath: workflow.outputFolder || '',
-				onFolderChange: async (folderPath: string) => {
-					workflow.outputFolder = folderPath;
+				placeholder: 'Search for folder...',
+				initialPath: workflow.outputFolder || '',
+				onChange: async (path: string) => {
+					workflow.outputFolder = path;
 					await plugin.saveSettings();
 				}
 			});
@@ -532,15 +525,14 @@ function displayTranscriptionWorkflowSettings(
 		// Output folder (only show if output type is new-note)
 		if (workflow.outputType === 'new-note') {
 			createPathPicker({
-				mode: 'folder-only',
 				containerEl: contentContainer,
 				app: plugin.app,
 				name: 'Output folder',
 				description: 'Folder where transcription notes will be created (leave empty to use default)',
-				folderPlaceholder: 'Default folder',
-				initialFolderPath: workflow.outputFolder || '',
-				onFolderChange: async (folderPath: string) => {
-					workflow.outputFolder = folderPath;
+				placeholder: 'Search for folder...',
+				initialPath: workflow.outputFolder || '',
+				onChange: async (path: string) => {
+					workflow.outputFolder = path;
 					await plugin.saveSettings();
 				}
 			});

From 07f900771cd2edf06760e87be91c4a6efb54f329 Mon Sep 17 00:00:00 2001
From: dalinicus <contact@dalinicus.com>
Date: Sun, 11 Jan 2026 20:34:12 -0600
Subject: [PATCH 5/7] add log pain

---
 .../output/at-cursor-output-handler.ts        |   9 +-
 .../output/new-note-output-handler.ts         |   5 +-
 src/handlers/output/popup-output-handler.ts   |   5 +-
 src/logging/index.ts                          |  10 ++
 src/logging/log-manager.ts                    | 124 ++++++++++++++++++
 src/logging/log-pane-view.ts                  | 118 +++++++++++++++++
 src/logging/types.ts                          |  52 ++++++++
 src/main.ts                                   |  49 ++++++-
 src/processing/video-processor.ts             |  20 ++-
 src/processing/workflow-executor.ts           |  53 ++++----
 styles.css                                    |  96 ++++++++++++++
 11 files changed, 494 insertions(+), 47 deletions(-)
 create mode 100644 src/logging/index.ts
 create mode 100644 src/logging/log-manager.ts
 create mode 100644 src/logging/log-pane-view.ts
 create mode 100644 src/logging/types.ts

diff --git a/src/handlers/output/at-cursor-output-handler.ts b/src/handlers/output/at-cursor-output-handler.ts
index 6d6347f..fbc8b88 100644
--- a/src/handlers/output/at-cursor-output-handler.ts
+++ b/src/handlers/output/at-cursor-output-handler.ts
@@ -1,5 +1,6 @@
-import { MarkdownView, Notice } from 'obsidian';
+import { MarkdownView } from 'obsidian';
 import { OutputHandler, OutputContext } from './types';
+import { logNotice, LogCategory } from '../../logging';
 
 /**
  * Output handler that inserts the AI response at the current cursor position
@@ -9,7 +10,7 @@ export class AtCursorOutputHandler implements OutputHandler {
     async handleOutput(responseText: string, context: OutputContext): Promise<void> {
         const activeView = context.app.workspace.getActiveViewOfType(MarkdownView);
         if (!activeView) {
-            new Notice('No active editor. Please open a note first.');
+            logNotice(LogCategory.WORKFLOW, 'No active editor. Please open a note first.');
             return;
         }
 
@@ -20,12 +21,12 @@ export class AtCursorOutputHandler implements OutputHandler {
             const from = editor.getCursor('from');
             editor.replaceSelection(responseText);
             this.moveCursorToEndOfText(editor, from, responseText);
-            new Notice('Response replaced selection');
+            logNotice(LogCategory.WORKFLOW, 'Response replaced selection');
         } else {
             const cursor = editor.getCursor();
             editor.replaceRange(responseText, cursor);
             this.moveCursorToEndOfText(editor, cursor, responseText);
-            new Notice('Response inserted at cursor');
+            logNotice(LogCategory.WORKFLOW, 'Response inserted at cursor');
         }
     }
 
diff --git a/src/handlers/output/new-note-output-handler.ts b/src/handlers/output/new-note-output-handler.ts
index 6eb33a5..26dde0d 100644
--- a/src/handlers/output/new-note-output-handler.ts
+++ b/src/handlers/output/new-note-output-handler.ts
@@ -1,6 +1,7 @@
-import { Notice, TFile } from 'obsidian';
+import { TFile } from 'obsidian';
 import { OutputHandler, OutputContext } from './types';
 import { generateFilenameTimestamp } from '../../utils/date-utils';
+import { logNotice, LogCategory } from '../../logging';
 
 /**
  * Output handler that creates a new note containing the AI response.
@@ -57,7 +58,7 @@ export class NewNoteOutputHandler implements OutputHandler {
             await leaf.openFile(file);
         }
 
-        new Notice(`Created note: ${file.name}`);
+        logNotice(LogCategory.WORKFLOW, `Created note: ${file.name}`);
     }
 }
 
diff --git a/src/handlers/output/popup-output-handler.ts b/src/handlers/output/popup-output-handler.ts
index 6dfeb3b..ca69a89 100644
--- a/src/handlers/output/popup-output-handler.ts
+++ b/src/handlers/output/popup-output-handler.ts
@@ -1,5 +1,6 @@
-import { App, Modal, Notice } from 'obsidian';
+import { App, Modal } from 'obsidian';
 import { OutputHandler, OutputContext } from './types';
+import { logNotice, LogCategory } from '../../logging';
 
 /**
  * Modal to display the AI response from a workflow execution.
@@ -30,7 +31,7 @@ export class WorkflowResultModal extends Modal {
         const copyButton = buttonContainer.createEl('button', { text: 'Copy to clipboard' });
         copyButton.addEventListener('click', () => {
             void navigator.clipboard.writeText(this.response).then(() => {
-                new Notice('Response copied to clipboard');
+                logNotice(LogCategory.WORKFLOW, 'Response copied to clipboard');
             });
         });
 
diff --git a/src/logging/index.ts b/src/logging/index.ts
new file mode 100644
index 0000000..d9eac2b
--- /dev/null
+++ b/src/logging/index.ts
@@ -0,0 +1,10 @@
+// Re-export types
+export { LogLevel, LogCategory } from './types';
+export type { LogEntry, LogListener, LogUnsubscribe, LogCategoryType } from './types';
+
+// Re-export log manager and convenience functions
+export { logManager, logDebug, logInfo, logWarn, logError, logNotice } from './log-manager';
+
+// Re-export log pane view
+export { VIEW_TYPE_LOG, LogPaneView } from './log-pane-view';
+
diff --git a/src/logging/log-manager.ts b/src/logging/log-manager.ts
new file mode 100644
index 0000000..eeb96dc
--- /dev/null
+++ b/src/logging/log-manager.ts
@@ -0,0 +1,124 @@
+import { Notice } from 'obsidian';
+import { LogEntry, LogLevel, LogListener, LogUnsubscribe, LogCategoryType } from './types';
+
+const MAX_LOGS = 500;
+
+/**
+ * Centralized log manager with subscription system for real-time log updates.
+ */
+class LogManager {
+    private logs: LogEntry[] = [];
+    private listeners: LogListener[] = [];
+
+    /**
+     * Add a log entry with the specified level, category, and message.
+     */
+    log(level: LogLevel, category: string, message: string, details?: unknown): void {
+        const entry: LogEntry = {
+            timestamp: new Date(),
+            level,
+            category,
+            message,
+            details
+        };
+
+        this.logs.push(entry);
+        if (this.logs.length > MAX_LOGS) {
+            this.logs = this.logs.slice(-MAX_LOGS);
+        }
+
+        // Console output based on level
+        const formattedMsg = `[AI Toolbox] [${category}] ${message}`;
+        switch (level) {
+            case LogLevel.ERROR:
+                console.error(formattedMsg, details !== undefined ? details : '');
+                break;
+            case LogLevel.WARN:
+                console.warn(formattedMsg, details !== undefined ? details : '');
+                break;
+            case LogLevel.DEBUG:
+                console.debug(formattedMsg, details !== undefined ? details : '');
+                break;
+            default:
+                console.log(formattedMsg, details !== undefined ? details : '');
+        }
+
+        this.notifyListeners();
+    }
+
+    /**
+     * Subscribe to log updates. Returns an unsubscribe function.
+     */
+    subscribe(callback: LogListener): LogUnsubscribe {
+        this.listeners.push(callback);
+        callback([...this.logs]); // Initial call with current logs
+        return () => {
+            this.listeners = this.listeners.filter(l => l !== callback);
+        };
+    }
+
+    private notifyListeners(): void {
+        const logsCopy = [...this.logs];
+        this.listeners.forEach(cb => cb(logsCopy));
+    }
+
+    /**
+     * Get a copy of all current logs.
+     */
+    getLogs(): LogEntry[] {
+        return [...this.logs];
+    }
+
+    /**
+     * Clear all logs.
+     */
+    clear(): void {
+        this.logs = [];
+        this.notifyListeners();
+    }
+
+    /**
+     * Get the count of logs at each level.
+     */
+    getLogCounts(): Record<LogLevel, number> {
+        const counts = {
+            [LogLevel.DEBUG]: 0,
+            [LogLevel.INFO]: 0,
+            [LogLevel.WARN]: 0,
+            [LogLevel.ERROR]: 0
+        };
+        for (const log of this.logs) {
+            counts[log.level]++;
+        }
+        return counts;
+    }
+}
+
+/** Singleton log manager instance */
+export const logManager = new LogManager();
+
+// Convenience functions for logging at specific levels
+export function logDebug(category: LogCategoryType | string, message: string, details?: unknown): void {
+    logManager.log(LogLevel.DEBUG, category, message, details);
+}
+
+export function logInfo(category: LogCategoryType | string, message: string, details?: unknown): void {
+    logManager.log(LogLevel.INFO, category, message, details);
+}
+
+export function logWarn(category: LogCategoryType | string, message: string, details?: unknown): void {
+    logManager.log(LogLevel.WARN, category, message, details);
+}
+
+export function logError(category: LogCategoryType | string, message: string, details?: unknown): void {
+    logManager.log(LogLevel.ERROR, category, message, details);
+}
+
+/**
+ * Log at INFO level and also display an Obsidian Notice to the user.
+ */
+export function logNotice(category: LogCategoryType | string, message: string, details?: unknown): void {
+    logManager.log(LogLevel.INFO, category, message, details);
+    new Notice(message);
+}
+
diff --git a/src/logging/log-pane-view.ts b/src/logging/log-pane-view.ts
new file mode 100644
index 0000000..ae612f0
--- /dev/null
+++ b/src/logging/log-pane-view.ts
@@ -0,0 +1,118 @@
+import { ItemView, WorkspaceLeaf } from 'obsidian';
+import { logManager } from './log-manager';
+import { LogEntry, LogLevel, LogUnsubscribe } from './types';
+import type AIToolboxPlugin from '../main';
+
+export const VIEW_TYPE_LOG = 'ai-toolbox-log';
+
+/**
+ * Log pane view for displaying AI Toolbox plugin logs in a dedicated Obsidian panel.
+ */
+export class LogPaneView extends ItemView {
+    plugin: AIToolboxPlugin;
+    private logContainer: HTMLElement;
+    private unsubscribe: LogUnsubscribe | null = null;
+    private autoScroll = true;
+    private filterLevel: LogLevel = LogLevel.DEBUG;
+
+    constructor(leaf: WorkspaceLeaf, plugin: AIToolboxPlugin) {
+        super(leaf);
+        this.plugin = plugin;
+    }
+
+    getViewType(): string {
+        return VIEW_TYPE_LOG;
+    }
+
+    getDisplayText(): string {
+        return 'AI Toolbox Log';
+    }
+
+    getIcon(): string {
+        return 'scroll-text';
+    }
+
+    async onOpen(): Promise<void> {
+        const container = this.containerEl.children[1] as HTMLElement;
+        container.empty();
+        container.addClass('ai-toolbox-log-view');
+
+        this.buildControls(container);
+        this.logContainer = container.createDiv({ cls: 'log-container' });
+
+        this.unsubscribe = logManager.subscribe(logs => this.renderLogs(logs));
+    }
+
+    private buildControls(container: HTMLElement): void {
+        const controls = container.createDiv({ cls: 'log-controls' });
+
+        // Filter dropdown
+        const filterLabel = controls.createEl('label', { cls: 'log-control-item' });
+        filterLabel.createSpan({ text: 'Level: ' });
+        const select = filterLabel.createEl('select', { cls: 'dropdown' });
+        const levels = ['DEBUG', 'INFO', 'WARN', 'ERROR'];
+        levels.forEach((level, i) => {
+            const option = select.createEl('option', { value: String(i), text: level });
+            if (i === this.filterLevel) option.selected = true;
+        });
+        select.addEventListener('change', () => {
+            this.filterLevel = parseInt(select.value);
+            this.renderLogs(logManager.getLogs());
+        });
+
+        // Auto-scroll toggle
+        const autoScrollLabel = controls.createEl('label', { cls: 'log-control-item' });
+        const checkbox = autoScrollLabel.createEl('input', { type: 'checkbox' });
+        checkbox.checked = this.autoScroll;
+        checkbox.addEventListener('change', () => {
+            this.autoScroll = checkbox.checked;
+        });
+        autoScrollLabel.appendText(' Auto-scroll');
+
+        // Spacer
+        controls.createDiv({ cls: 'log-controls-spacer' });
+
+        // Clear button
+        const clearBtn = controls.createEl('button', { text: 'Clear', cls: 'mod-warning' });
+        clearBtn.addEventListener('click', () => logManager.clear());
+
+        // Copy button
+        const copyBtn = controls.createEl('button', { text: 'Copy All' });
+        copyBtn.addEventListener('click', () => this.copyLogs());
+    }
+
+    private renderLogs(logs: LogEntry[]): void {
+        this.logContainer.empty();
+        const filtered = logs.filter(l => l.level >= this.filterLevel);
+
+        for (const entry of filtered) {
+            const levelName = LogLevel[entry.level].toLowerCase();
+            const line = this.logContainer.createDiv({ cls: `log-entry log-${levelName}` });
+
+            const time = entry.timestamp.toLocaleTimeString();
+            line.createSpan({ cls: 'log-time', text: `[${time}]` });
+            line.createSpan({ cls: 'log-level', text: ` [${LogLevel[entry.level]}]` });
+            line.createSpan({ cls: 'log-category', text: ` [${entry.category}]` });
+            line.createSpan({ cls: 'log-message', text: ` ${entry.message}` });
+        }
+
+        if (this.autoScroll) {
+            this.logContainer.scrollTop = this.logContainer.scrollHeight;
+        }
+    }
+
+    private copyLogs(): void {
+        const text = logManager.getLogs()
+            .map(e => `[${e.timestamp.toISOString()}] [${LogLevel[e.level]}] [${e.category}] ${e.message}`)
+            .join('\n');
+        navigator.clipboard.writeText(text);
+    }
+
+    async onClose(): Promise<void> {
+        if (this.unsubscribe) {
+            this.unsubscribe();
+            this.unsubscribe = null;
+        }
+    }
+}
+
diff --git a/src/logging/types.ts b/src/logging/types.ts
new file mode 100644
index 0000000..776105a
--- /dev/null
+++ b/src/logging/types.ts
@@ -0,0 +1,52 @@
+/**
+ * Log level enumeration for categorizing log messages.
+ */
+export enum LogLevel {
+    DEBUG = 0,
+    INFO = 1,
+    WARN = 2,
+    ERROR = 3
+}
+
+/**
+ * Represents a single log entry with timestamp, level, category, and message.
+ */
+export interface LogEntry {
+    /** When the log entry was created */
+    timestamp: Date;
+    /** Severity level of the log */
+    level: LogLevel;
+    /** Category/module that generated the log (e.g., "workflow", "provider", "transcription") */
+    category: string;
+    /** The log message */
+    message: string;
+    /** Optional additional details (objects, errors, etc.) */
+    details?: unknown;
+}
+
+/**
+ * Callback type for log subscription listeners.
+ */
+export type LogListener = (logs: LogEntry[]) => void;
+
+/**
+ * Unsubscribe function returned when subscribing to log updates.
+ */
+export type LogUnsubscribe = () => void;
+
+/**
+ * Log category constants for consistent categorization.
+ */
+export const LogCategory = {
+    PLUGIN: 'plugin',
+    WORKFLOW: 'workflow',
+    PROVIDER: 'provider',
+    TRANSCRIPTION: 'transcription',
+    INPUT: 'input',
+    OUTPUT: 'output',
+    VIDEO: 'video',
+    AUDIO: 'audio'
+} as const;
+
+export type LogCategoryType = typeof LogCategory[keyof typeof LogCategory];
+
diff --git a/src/main.ts b/src/main.ts
index 4e3b00c..b9097cb 100644
--- a/src/main.ts
+++ b/src/main.ts
@@ -1,7 +1,8 @@
-import { Notice, Plugin } from 'obsidian';
+import { Plugin } from 'obsidian';
 import { DEFAULT_SETTINGS, AIToolboxSettings, AIToolboxSettingTab } from "./settings/index";
 import { WorkflowSuggesterModal } from "./components/workflow-suggester";
 import { executeWorkflow } from "./processing/workflow-executor";
+import { VIEW_TYPE_LOG, LogPaneView, logInfo, logNotice, LogCategory } from "./logging";
 
 export default class AIToolboxPlugin extends Plugin {
 	settings: AIToolboxSettings;
@@ -9,6 +10,14 @@ export default class AIToolboxPlugin extends Plugin {
 	async onload() {
 		await this.loadSettings();
 
+		// Register log view
+		this.registerView(VIEW_TYPE_LOG, (leaf) => new LogPaneView(leaf, this));
+
+		// Add ribbon icon for log view
+		this.addRibbonIcon('scroll-text', 'Show AI Toolbox Log', () => {
+			void this.activateLogView();
+		});
+
 		// Add command to execute custom workflows
 		this.addCommand({
 			id: 'execute-workflow',
@@ -16,8 +25,44 @@ export default class AIToolboxPlugin extends Plugin {
 			callback: () => this.showWorkflowSuggester()
 		});
 
+		// Add command to show log view
+		this.addCommand({
+			id: 'show-log',
+			name: 'Show log',
+			callback: () => {
+				void this.activateLogView();
+			}
+		});
+
 		// This adds a settings tab so the user can configure various aspects of the plugin
 		this.addSettingTab(new AIToolboxSettingTab(this.app, this));
+
+		logInfo(LogCategory.PLUGIN, 'AI Toolbox plugin loaded');
+	}
+
+	onunload() {
+		// Clean up log view leaves
+		this.app.workspace.detachLeavesOfType(VIEW_TYPE_LOG);
+	}
+
+	/**
+	 * Activate and reveal the log view in the right sidebar.
+	 */
+	private async activateLogView(): Promise<void> {
+		const { workspace } = this.app;
+		let leaf = workspace.getLeavesOfType(VIEW_TYPE_LOG)[0];
+
+		if (!leaf) {
+			const rightLeaf = workspace.getRightLeaf(false);
+			if (rightLeaf) {
+				leaf = rightLeaf;
+				await leaf.setViewState({ type: VIEW_TYPE_LOG, active: true });
+			}
+		}
+
+		if (leaf) {
+			workspace.revealLeaf(leaf);
+		}
 	}
 
 	/**
@@ -27,7 +72,7 @@ export default class AIToolboxPlugin extends Plugin {
 		const availableWorkflows = this.settings.workflows.filter(w => w.showInCommand);
 
 		if (availableWorkflows.length === 0) {
-			new Notice('No workflows available. You can configure workflows from the workflows tab in settings.');
+			logNotice(LogCategory.WORKFLOW, 'No workflows available. You can configure workflows from the workflows tab in settings.');
 			return;
 		}
 
diff --git a/src/processing/video-processor.ts b/src/processing/video-processor.ts
index 7e3e6f7..cd9ed24 100644
--- a/src/processing/video-processor.ts
+++ b/src/processing/video-processor.ts
@@ -3,9 +3,9 @@ import * as path from 'path';
 import * as os from 'os';
 import * as fs from 'fs';
 import { Buffer } from 'buffer';
-import { Notice } from 'obsidian';
 import { AIToolboxSettings } from '../settings';
 import { videoPlatformRegistry, VideoMetadata } from './video-platforms';
+import { logNotice, LogCategory } from '../logging';
 
 /**
  * Configuration for video processing operations.
@@ -219,18 +219,18 @@ function settingsToProcessorConfig(settings: AIToolboxSettings): VideoProcessorC
 export async function extractAudioFromUrl(url: string, settings: AIToolboxSettings): Promise<ExtractAudioResult | null> {
     try {
         if (!url || !url.trim()) {
-            new Notice('URL is empty');
+            logNotice(LogCategory.TRANSCRIPTION, 'URL is empty');
             return null;
         }
 
         const trimmedUrl = url.trim();
 
         if (!videoPlatformRegistry.isValidVideoUrl(trimmedUrl)) {
-            new Notice('The provided text is not a valid video URL');
+            logNotice(LogCategory.TRANSCRIPTION, 'The provided text is not a valid video URL');
             return null;
         }
 
-        new Notice('Preparing video for transcription...');
+        logNotice(LogCategory.TRANSCRIPTION, 'Preparing video for transcription...');
 
         const handler = videoPlatformRegistry.findHandlerForUrl(trimmedUrl);
         const filenameTemplate = handler
@@ -243,7 +243,7 @@ export async function extractAudioFromUrl(url: string, settings: AIToolboxSettin
 
         const ytdlpResult = await runYtDlp(trimmedUrl, outputTemplate, config);
 
-        new Notice(`Audio extracted successfully!\nReady for transcription.\nSaved to: ${path.dirname(ytdlpResult.audioFilePath)}`);
+        logNotice(LogCategory.TRANSCRIPTION, `Audio extracted successfully!\nReady for transcription.\nSaved to: ${path.dirname(ytdlpResult.audioFilePath)}`);
 
         return {
             audioFilePath: ytdlpResult.audioFilePath,
@@ -257,9 +257,7 @@ export async function extractAudioFromUrl(url: string, settings: AIToolboxSettin
         };
 
     } catch (error) {
-        console.error('Video audio extraction error:', error);
-        const errorMessage = error instanceof Error ? error.message : String(error);
-        new Notice(`Failed to extract audio for transcription: ${errorMessage}`);
+        logNotice(LogCategory.TRANSCRIPTION, `Failed to extract audio for transcription: ${error instanceof Error ? error.message : String(error)}`, error);
         return null;
     }
 }
@@ -275,16 +273,14 @@ export async function extractAudioFromClipboard(settings: AIToolboxSettings): Pr
         const clipboardText = await navigator.clipboard.readText();
 
         if (!clipboardText) {
-            new Notice('Clipboard is empty');
+            logNotice(LogCategory.TRANSCRIPTION, 'Clipboard is empty');
             return null;
         }
 
         return extractAudioFromUrl(clipboardText, settings);
 
     } catch (error) {
-        console.error('Clipboard read error:', error);
-        const errorMessage = error instanceof Error ? error.message : String(error);
-        new Notice(`Failed to read clipboard: ${errorMessage}`);
+        logNotice(LogCategory.TRANSCRIPTION, `Failed to read clipboard: ${error instanceof Error ? error.message : String(error)}`, error);
         return null;
     }
 }
diff --git a/src/processing/workflow-executor.ts b/src/processing/workflow-executor.ts
index af5c6f8..0fae04f 100644
--- a/src/processing/workflow-executor.ts
+++ b/src/processing/workflow-executor.ts
@@ -1,4 +1,4 @@
-import { App, Notice, TFile } from 'obsidian';
+import { App, TFile } from 'obsidian';
 import { WorkflowConfig, AIToolboxSettings, TranscriptionSourceType } from '../settings';
 import { createWorkflowProvider, ChatMessage, TranscriptionOptions, TranscriptionResult } from '../providers';
 import { videoPlatformRegistry } from './video-platforms';
@@ -30,6 +30,7 @@ import {
     hasContextTokens,
     formatTranscriptionWithTimestamps
 } from './workflow-chaining';
+import { logInfo, logDebug, logWarn, logNotice, LogCategory } from '../logging';
 
 /**
  * Create an output handler based on the workflow's output type.
@@ -101,21 +102,20 @@ async function getPromptText(app: App, workflow: WorkflowConfig): Promise<string
 	if (sourceType === 'from-file') {
 		const filePath = workflow.promptFilePath;
 		if (!filePath || !filePath.trim()) {
-			new Notice(`Workflow "${workflow.name}" has no prompt file configured. Please select a file in settings.`);
+			logNotice(LogCategory.WORKFLOW, `Workflow "${workflow.name}" has no prompt file configured. Please select a file in settings.`);
 			return null;
 		}
 
 		const file = app.vault.getAbstractFileByPath(filePath);
 		if (!file || !(file instanceof TFile)) {
-			new Notice(`Prompt file "${filePath}" not found in vault for workflow "${workflow.name}".`);
+			logNotice(LogCategory.WORKFLOW, `Prompt file "${filePath}" not found in vault for workflow "${workflow.name}".`);
 			return null;
 		}
 
 		try {
 			return await app.vault.read(file);
 		} catch (error) {
-			console.error('Error reading prompt file:', error);
-			new Notice(`Failed to read prompt file "${filePath}": ${error instanceof Error ? error.message : String(error)}`);
+			logNotice(LogCategory.WORKFLOW, `Failed to read prompt file "${filePath}": ${error instanceof Error ? error.message : String(error)}`, error);
 			return null;
 		}
 	}
@@ -146,13 +146,13 @@ async function executeDependencies(
 		// Check for circular dependency (already in execution stack)
 		if (executionStack.has(depId)) {
 			const depWorkflow = settings.workflows.find(w => w.id === depId);
-			new Notice(`Circular dependency detected: "${depWorkflow?.name ?? depId}" is already being executed.`);
+			logNotice(LogCategory.WORKFLOW, `Circular dependency detected: "${depWorkflow?.name ?? depId}" is already being executed.`);
 			return false;
 		}
 
 		const depWorkflow = settings.workflows.find(w => w.id === depId);
 		if (!depWorkflow) {
-			new Notice(`Dependency workflow not found: ${depId}`);
+			logNotice(LogCategory.WORKFLOW, `Dependency workflow not found: ${depId}`);
 			return false;
 		}
 
@@ -168,11 +168,11 @@ async function executeDependencies(
 		}
 
 		// Execute the dependency workflow
-		new Notice(`Executing dependency: ${depWorkflow.name}...`);
+		logNotice(LogCategory.WORKFLOW, `Executing dependency: ${depWorkflow.name}...`);
 		const result = await executeWorkflowInternal(app, settings, depWorkflow, results);
 
 		if (!result.success) {
-			new Notice(`Dependency workflow "${depWorkflow.name}" failed: ${result.error}`);
+			logNotice(LogCategory.WORKFLOW, `Dependency workflow "${depWorkflow.name}" failed: ${result.error}`);
 			return false;
 		}
 
@@ -197,11 +197,13 @@ export async function executeWorkflow(
 	settings: AIToolboxSettings,
 	workflow: WorkflowConfig
 ): Promise<void> {
+	logInfo(LogCategory.WORKFLOW, `Starting workflow: ${workflow.name}`, { type: workflow.type, id: workflow.id });
+
 	// Check for circular dependencies before starting
 	if (hasWorkflowDependencies(workflow)) {
 		const cycle = detectCircularDependency(workflow, settings);
 		if (cycle.length > 0) {
-			new Notice(`Circular dependency detected: ${cycle.join(' → ')}`);
+			logNotice(LogCategory.WORKFLOW, `Circular dependency detected: ${cycle.join(' → ')}`);
 			return;
 		}
 
@@ -266,7 +268,7 @@ async function executeChatWorkflow(
 ): Promise<void> {
 	// Validate workflow has a provider configured
 	if (!workflow.provider) {
-		new Notice(`Workflow "${workflow.name}" has no provider configured. Please configure a provider in settings.`);
+		logNotice(LogCategory.WORKFLOW, `Workflow "${workflow.name}" has no provider configured. Please configure a provider in settings.`);
 		return;
 	}
 
@@ -289,31 +291,33 @@ async function executeChatWorkflow(
 
 	// Validate workflow has text
 	if (!promptText.trim()) {
-		new Notice(`Workflow "${workflow.name}" has no prompt text. Please add prompt text in settings.`);
+		logNotice(LogCategory.WORKFLOW, `Workflow "${workflow.name}" has no prompt text. Please add prompt text in settings.`);
 		return;
 	}
 
 	// Create the provider
 	const provider = createWorkflowProvider(settings, workflow);
 	if (!provider) {
-		new Notice(`Could not find the configured provider for workflow "${workflow.name}". Please check your settings.`);
+		logNotice(LogCategory.WORKFLOW, `Could not find the configured provider for workflow "${workflow.name}". Please check your settings.`);
 		return;
 	}
 
 	// Validate provider supports chat
 	if (!provider.supportsChat()) {
-		new Notice(`The provider for workflow "${workflow.name}" does not support chat. Please select a chat-capable model.`);
+		logNotice(LogCategory.WORKFLOW, `The provider for workflow "${workflow.name}" does not support chat. Please select a chat-capable model.`);
 		return;
 	}
 
 	try {
-		new Notice(`Executing workflow: ${workflow.name}...`);
+		logNotice(LogCategory.WORKFLOW, `Executing workflow: ${workflow.name}...`);
+		logDebug(LogCategory.WORKFLOW, `Sending chat request to provider`);
 
 		const messages: ChatMessage[] = [
 			{ role: 'user', content: promptText }
 		];
 
 		const result = await provider.sendChat(messages);
+		logInfo(LogCategory.WORKFLOW, `Chat workflow completed: ${workflow.name}`);
 
 		const outputType = workflow.outputType || 'popup';
 		const handler = createOutputHandler(outputType);
@@ -326,9 +330,7 @@ async function executeChatWorkflow(
 		await handler.handleOutput(result.content, context);
 
 	} catch (error) {
-		console.error('Workflow execution error:', error);
-		const errorMessage = error instanceof Error ? error.message : String(error);
-		new Notice(`Failed to execute workflow: ${errorMessage}`);
+		logNotice(LogCategory.WORKFLOW, `Workflow execution error: ${workflow.name}`, error);
 	}
 }
 
@@ -410,18 +412,18 @@ async function executeTranscriptionWorkflow(
 	_dependencyResults?: WorkflowResultsMap
 ): Promise<void> {
 	if (!workflow.provider) {
-		new Notice(`Workflow "${workflow.name}" has no provider configured. Please configure a provider in settings.`);
+		logNotice(LogCategory.TRANSCRIPTION, `Workflow "${workflow.name}" has no provider configured. Please configure a provider in settings.`);
 		return;
 	}
 
 	const provider = createWorkflowProvider(settings, workflow);
 	if (!provider) {
-		new Notice(`Could not find the configured provider for workflow "${workflow.name}". Please check your settings.`);
+		logNotice(LogCategory.TRANSCRIPTION, `Could not find the configured provider for workflow "${workflow.name}". Please check your settings.`);
 		return;
 	}
 
 	if (!provider.supportsTranscription()) {
-		new Notice(`The provider for workflow "${workflow.name}" does not support transcription. Please select a transcription-capable model.`);
+		logNotice(LogCategory.TRANSCRIPTION, `The provider for workflow "${workflow.name}" does not support transcription. Please select a transcription-capable model.`);
 		return;
 	}
 
@@ -435,7 +437,8 @@ async function executeTranscriptionWorkflow(
 			return;
 		}
 
-		new Notice(`Transcribing audio...`);
+		logNotice(LogCategory.TRANSCRIPTION, `Transcribing audio...`);
+		logDebug(LogCategory.TRANSCRIPTION, `Transcribing audio file: ${inputResult.audioFilePath}`);
 
 		const transcriptionOptions: TranscriptionOptions = {
 			timestampGranularity: workflow.timestampGranularity ?? 'disabled',
@@ -443,6 +446,8 @@ async function executeTranscriptionWorkflow(
 		};
 
 		const transcriptionResult = await provider.transcribeAudio(inputResult.audioFilePath, transcriptionOptions);
+		logInfo(LogCategory.TRANSCRIPTION, `Transcription completed: ${workflow.name}`);
+
 		// Use timestamped version for display output if available, otherwise plain text
 		const formattedText = transcriptionResult.chunks.length > 0
 			? formatTranscriptionWithTimestamps(transcriptionResult.chunks)
@@ -456,9 +461,7 @@ async function executeTranscriptionWorkflow(
 		await handler.handleOutput(formattedText, context);
 
 	} catch (error) {
-		console.error('Transcription workflow execution error:', error);
-		const errorMessage = error instanceof Error ? error.message : String(error);
-		new Notice(`Failed to transcribe audio: ${errorMessage}`);
+		logNotice(LogCategory.TRANSCRIPTION, `Transcription workflow error: ${workflow.name}`, error);
 	}
 }
 
diff --git a/styles.css b/styles.css
index 8912930..d1d0cc8 100644
--- a/styles.css
+++ b/styles.css
@@ -635,3 +635,99 @@ button.mod-success:hover {
 .additional-settings-heading .setting-item-name {
 	font-size: 1.17em;
 }
+
+/* ==================== Log View Styles ==================== */
+
+.ai-toolbox-log-view {
+	display: flex;
+	flex-direction: column;
+	height: 100%;
+	overflow: hidden;
+}
+
+.log-controls {
+	display: flex;
+	align-items: center;
+	gap: 10px;
+	padding: 8px 12px;
+	border-bottom: 1px solid var(--background-modifier-border);
+	flex-wrap: wrap;
+	flex-shrink: 0;
+}
+
+.log-control-item {
+	display: flex;
+	align-items: center;
+	gap: 4px;
+	font-size: 13px;
+}
+
+.log-controls-spacer {
+	flex: 1;
+}
+
+.log-container {
+	flex: 1;
+	overflow-y: auto;
+	font-family: var(--font-monospace);
+	font-size: 12px;
+	padding: 8px 12px;
+	user-select: text;
+	-webkit-user-select: text;
+}
+
+.log-entry {
+	padding: 3px 0;
+	border-bottom: 1px solid var(--background-modifier-border-hover);
+	line-height: 1.4;
+	word-break: break-word;
+}
+
+.log-entry:last-child {
+	border-bottom: none;
+}
+
+.log-time {
+	color: var(--text-muted);
+}
+
+.log-level {
+	font-weight: 600;
+}
+
+.log-category {
+	color: var(--text-accent);
+}
+
+.log-message {
+	color: var(--text-normal);
+}
+
+/* Log level specific styles */
+.log-debug {
+	opacity: 0.7;
+}
+
+.log-debug .log-level {
+	color: var(--text-muted);
+}
+
+.log-info .log-level {
+	color: var(--text-accent);
+}
+
+.log-warn {
+	background-color: rgba(255, 200, 0, 0.08);
+}
+
+.log-warn .log-level {
+	color: var(--text-warning);
+}
+
+.log-error {
+	background-color: rgba(255, 100, 100, 0.08);
+}
+
+.log-error .log-level {
+	color: var(--text-error);
+}

From 229e1cec4adc09a9dc3a57b85289cbdac7628f48 Mon Sep 17 00:00:00 2001
From: dalinicus <contact@dalinicus.com>
Date: Sun, 11 Jan 2026 21:21:31 -0600
Subject: [PATCH 6/7] actions

---
 src/main.ts                         |   2 +-
 src/processing/action-executor.ts   | 299 ++++++++++
 src/processing/workflow-chaining.ts |  61 +-
 src/processing/workflow-executor.ts | 724 ++++++++---------------
 src/providers/index.ts              |   1 +
 src/providers/provider-factory.ts   |  23 +-
 src/settings/index.ts               |  16 +-
 src/settings/types.ts               | 107 +++-
 src/settings/workflows.ts           | 865 +++++++++-------------------
 src/tokens/types.ts                 |  48 +-
 styles.css                          |  27 +
 11 files changed, 1021 insertions(+), 1152 deletions(-)
 create mode 100644 src/processing/action-executor.ts

diff --git a/src/main.ts b/src/main.ts
index b9097cb..b0b7883 100644
--- a/src/main.ts
+++ b/src/main.ts
@@ -69,7 +69,7 @@ export default class AIToolboxPlugin extends Plugin {
 	 * Show the workflow suggester modal and execute the selected workflow.
 	 */
 	private showWorkflowSuggester(): void {
-		const availableWorkflows = this.settings.workflows.filter(w => w.showInCommand);
+		const availableWorkflows = this.settings.workflows;
 
 		if (availableWorkflows.length === 0) {
 			logNotice(LogCategory.WORKFLOW, 'No workflows available. You can configure workflows from the workflows tab in settings.');
diff --git a/src/processing/action-executor.ts b/src/processing/action-executor.ts
new file mode 100644
index 0000000..b212751
--- /dev/null
+++ b/src/processing/action-executor.ts
@@ -0,0 +1,299 @@
+import { App, TFile } from 'obsidian';
+import { AIToolboxSettings, ChatAction, TranscriptionAction, WorkflowAction, TranscriptionSourceType } from '../settings';
+import { createActionProvider, ChatMessage, TranscriptionOptions } from '../providers';
+import {
+    InputHandler,
+    InputContext,
+    InputResult,
+    VaultFileInputHandler,
+    ClipboardUrlInputHandler,
+    SelectionUrlInputHandler
+} from '../handlers';
+import {
+    createChatWorkflowTokens,
+    createTranscriptionWorkflowTokens,
+    gatherContextValues,
+    replaceContextTokens,
+    hasContextTokens
+} from './workflow-chaining';
+import { logInfo, logDebug, logNotice, LogCategory } from '../logging';
+
+/**
+ * Result from executing a single action
+ */
+export interface ActionExecutionResult {
+    /** The action that was executed */
+    actionId: string;
+    /** The action type */
+    actionType: 'chat' | 'transcription';
+    /** Whether execution succeeded */
+    success: boolean;
+    /** Error message if failed */
+    error?: string;
+    /** Token values produced by this action */
+    tokens: Record<string, string>;
+    /** Additional metadata (e.g., for transcription note title) */
+    metadata?: {
+        noteTitle?: string;
+        inputResult?: InputResult;
+    };
+}
+
+/**
+ * Map of action ID to execution results
+ */
+export type ActionResultsMap = Map<string, ActionExecutionResult>;
+
+/**
+ * Context for action execution
+ */
+export interface ActionExecutionContext {
+    app: App;
+    settings: AIToolboxSettings;
+    /** Results from previously executed actions in this workflow */
+    previousResults: ActionResultsMap;
+    /** Results from dependency workflows */
+    dependencyResults: Map<string, ActionExecutionResult>;
+    /** The workflow name (for logging) */
+    workflowName: string;
+}
+
+/**
+ * Replace action tokens in text with values from previous action results.
+ * Tokens are in the format {{actionId.tokenName}}
+ */
+export function replaceActionTokens(
+    text: string,
+    results: ActionResultsMap
+): string {
+    const tokenPattern = /\{\{([a-zA-Z0-9_-]+)\.([a-zA-Z0-9_]+)\}\}/g;
+
+    return text.replace(tokenPattern, (match, actionId: string, tokenName: string) => {
+        const result = results.get(actionId);
+        if (!result) {
+            return match;
+        }
+
+        const tokenValue = result.tokens[tokenName];
+        if (tokenValue === undefined) {
+            return match;
+        }
+
+        return tokenValue;
+    });
+}
+
+/**
+ * Create an input handler based on the transcription source type.
+ */
+function createInputHandler(sourceType: TranscriptionSourceType): InputHandler {
+    switch (sourceType) {
+        case 'select-file-from-vault':
+            return new VaultFileInputHandler();
+        case 'url-from-clipboard':
+            return new ClipboardUrlInputHandler();
+        case 'url-from-selection':
+            return new SelectionUrlInputHandler();
+        default:
+            return new VaultFileInputHandler();
+    }
+}
+
+/**
+ * Get the prompt text for a chat action, loading from file if needed.
+ */
+async function getPromptText(app: App, action: ChatAction): Promise<string | null> {
+    const sourceType = action.promptSourceType ?? 'inline';
+
+    if (sourceType === 'from-file') {
+        const filePath = action.promptFilePath;
+        if (!filePath || !filePath.trim()) {
+            logNotice(LogCategory.WORKFLOW, `Action "${action.name}" has no prompt file configured.`);
+            return null;
+        }
+
+        const file = app.vault.getAbstractFileByPath(filePath);
+        if (!file || !(file instanceof TFile)) {
+            logNotice(LogCategory.WORKFLOW, `Prompt file "${filePath}" not found for action "${action.name}".`);
+            return null;
+        }
+
+        try {
+            return await app.vault.read(file);
+        } catch (error) {
+            logNotice(LogCategory.WORKFLOW, `Failed to read prompt file: ${error instanceof Error ? error.message : String(error)}`);
+            return null;
+        }
+    }
+
+    return action.promptText;
+}
+
+/**
+ * Execute a chat action and return the result.
+ */
+export async function executeChatAction(
+    action: ChatAction,
+    context: ActionExecutionContext
+): Promise<ActionExecutionResult> {
+    const baseResult: ActionExecutionResult = {
+        actionId: action.id,
+        actionType: 'chat',
+        success: false,
+        tokens: {}
+    };
+
+    if (!action.provider) {
+        return { ...baseResult, error: 'No provider configured' };
+    }
+
+    const provider = createActionProvider(context.settings, action);
+    if (!provider) {
+        return { ...baseResult, error: 'Provider not found' };
+    }
+
+    if (!provider.supportsChat()) {
+        return { ...baseResult, error: 'Provider does not support chat' };
+    }
+
+    let promptText = await getPromptText(context.app, action);
+    if (promptText === null) {
+        return { ...baseResult, error: 'Failed to load prompt text' };
+    }
+
+    // Replace tokens from previous actions
+    if (context.previousResults.size > 0) {
+        promptText = replaceActionTokens(promptText, context.previousResults);
+    }
+
+    // Replace tokens from dependency workflows
+    if (context.dependencyResults.size > 0) {
+        promptText = replaceActionTokens(promptText, context.dependencyResults);
+    }
+
+    // Replace context tokens ({{selection}}, {{clipboard}}, etc.)
+    if (hasContextTokens(promptText)) {
+        const contextValues = await gatherContextValues(context.app);
+        promptText = replaceContextTokens(promptText, contextValues);
+    }
+
+    if (!promptText.trim()) {
+        return { ...baseResult, error: 'Empty prompt text' };
+    }
+
+    try {
+        logDebug(LogCategory.WORKFLOW, `Executing chat action: ${action.name}`);
+
+        const messages: ChatMessage[] = [
+            { role: 'user', content: promptText }
+        ];
+
+        const result = await provider.sendChat(messages);
+        logInfo(LogCategory.WORKFLOW, `Chat action completed: ${action.name}`);
+
+        return {
+            ...baseResult,
+            success: true,
+            tokens: createChatWorkflowTokens(promptText, result.content)
+        };
+    } catch (error) {
+        const errorMessage = error instanceof Error ? error.message : String(error);
+        return { ...baseResult, error: errorMessage };
+    }
+}
+
+/**
+ * Execute a transcription action and return the result.
+ */
+export async function executeTranscriptionAction(
+    action: TranscriptionAction,
+    context: ActionExecutionContext
+): Promise<ActionExecutionResult> {
+    const baseResult: ActionExecutionResult = {
+        actionId: action.id,
+        actionType: 'transcription',
+        success: false,
+        tokens: {}
+    };
+
+    if (!action.provider) {
+        return { ...baseResult, error: 'No provider configured' };
+    }
+
+    const provider = createActionProvider(context.settings, action);
+    if (!provider) {
+        return { ...baseResult, error: 'Provider not found' };
+    }
+
+    if (!provider.supportsTranscription()) {
+        return { ...baseResult, error: 'Provider does not support transcription' };
+    }
+
+    const sourceType: TranscriptionSourceType = action.transcriptionContext?.sourceType ?? 'url-from-clipboard';
+    const inputHandler = createInputHandler(sourceType);
+
+    // Create a minimal workflow-like object for InputContext compatibility
+    const inputContext: InputContext = {
+        app: context.app,
+        settings: context.settings,
+        workflow: {
+            id: action.id,
+            name: action.name,
+            actions: [],
+            outputType: 'popup',
+            outputFolder: ''
+        }
+    };
+
+    try {
+        logDebug(LogCategory.TRANSCRIPTION, `Executing transcription action: ${action.name}`);
+
+        const inputResult = await inputHandler.getInput(inputContext);
+        if (!inputResult) {
+            return { ...baseResult, error: 'No input provided or cancelled' };
+        }
+
+        logNotice(LogCategory.TRANSCRIPTION, `Transcribing audio...`);
+
+        const timestampGranularity = action.timestampGranularity ?? 'disabled';
+        const transcriptionOptions: TranscriptionOptions = {
+            timestampGranularity,
+            language: action.language || undefined
+        };
+
+        const transcriptionResult = await provider.transcribeAudio(inputResult.audioFilePath, transcriptionOptions);
+        logInfo(LogCategory.TRANSCRIPTION, `Transcription action completed: ${action.name}`);
+
+        const tokenMetadata = {
+            ...inputResult.metadata,
+            sourceUrl: inputResult.sourceUrl
+        };
+
+        return {
+            ...baseResult,
+            success: true,
+            tokens: createTranscriptionWorkflowTokens(transcriptionResult, tokenMetadata, timestampGranularity),
+            metadata: {
+                inputResult
+            }
+        };
+    } catch (error) {
+        const errorMessage = error instanceof Error ? error.message : String(error);
+        return { ...baseResult, error: errorMessage };
+    }
+}
+
+/**
+ * Execute any action type and return the result.
+ */
+export async function executeAction(
+    action: WorkflowAction,
+    context: ActionExecutionContext
+): Promise<ActionExecutionResult> {
+    if (action.type === 'chat') {
+        return executeChatAction(action, context);
+    } else {
+        return executeTranscriptionAction(action, context);
+    }
+}
+
diff --git a/src/processing/workflow-chaining.ts b/src/processing/workflow-chaining.ts
index c660f9b..a3de469 100644
--- a/src/processing/workflow-chaining.ts
+++ b/src/processing/workflow-chaining.ts
@@ -23,59 +23,27 @@ export interface WorkflowExecutionResult {
  */
 export type WorkflowResultsMap = Map<string, WorkflowExecutionResult>;
 
-/**
- * Context for tracking workflow dependency execution
- */
-export interface DependencyExecutionContext {
-    /** The Obsidian App instance */
-    app: App;
-    /** Plugin settings */
-    settings: AIToolboxSettings;
-    /** Results from executed dependencies */
-    results: WorkflowResultsMap;
-    /** Workflow IDs currently in the execution stack (for circular detection) */
-    executionStack: Set<string>;
-}
-
 /**
  * Detect circular dependencies in a workflow's dependency chain.
- * Returns an array of workflow names forming the cycle, or empty if no cycle.
+ * Currently workflows don't have cross-workflow dependencies (actions chain within a workflow).
+ * This function is kept for potential future use.
  */
 export function detectCircularDependency(
-    workflow: WorkflowConfig,
-    settings: AIToolboxSettings,
-    visited: Set<string> = new Set(),
-    path: string[] = []
+    _workflow: WorkflowConfig,
+    _settings: AIToolboxSettings,
+    _visited: Set<string> = new Set(),
+    _path: string[] = []
 ): string[] {
-    if (visited.has(workflow.id)) {
-        // Found a cycle - return the path from the first occurrence
-        const cycleStart = path.indexOf(workflow.name);
-        return [...path.slice(cycleStart), workflow.name];
-    }
-
-    visited.add(workflow.id);
-    path.push(workflow.name);
-
-    const workflowContexts = workflow.workflowContexts ?? [];
-    for (const ctx of workflowContexts) {
-        const depWorkflow = settings.workflows.find(w => w.id === ctx.workflowId);
-        if (depWorkflow) {
-            const cycle = detectCircularDependency(depWorkflow, settings, visited, path);
-            if (cycle.length > 0) {
-                return cycle;
-            }
-        }
-    }
-
-    path.pop();
+    // No cross-workflow dependencies in current design
     return [];
 }
 
 /**
- * Check if a workflow has any workflow dependencies configured
+ * Check if a workflow has any workflow dependencies configured.
+ * Currently workflows don't have cross-workflow dependencies.
  */
-export function hasWorkflowDependencies(workflow: WorkflowConfig): boolean {
-    return (workflow.workflowContexts?.length ?? 0) > 0;
+export function hasWorkflowDependencies(_workflow: WorkflowConfig): boolean {
+    return false;
 }
 
 /**
@@ -183,10 +151,11 @@ export function createTranscriptionWorkflowTokens(
 }
 
 /**
- * Get ordered list of dependency workflow IDs for a workflow
+ * Get ordered list of dependency workflow IDs for a workflow.
+ * Currently workflows don't have cross-workflow dependencies.
  */
-export function getDependencyWorkflowIds(workflow: WorkflowConfig): string[] {
-    return (workflow.workflowContexts ?? []).map(ctx => ctx.workflowId);
+export function getDependencyWorkflowIds(_workflow: WorkflowConfig): string[] {
+    return [];
 }
 
 /**
diff --git a/src/processing/workflow-executor.ts b/src/processing/workflow-executor.ts
index 0fae04f..a5878dd 100644
--- a/src/processing/workflow-executor.ts
+++ b/src/processing/workflow-executor.ts
@@ -1,6 +1,5 @@
-import { App, TFile } from 'obsidian';
-import { WorkflowConfig, AIToolboxSettings, TranscriptionSourceType } from '../settings';
-import { createWorkflowProvider, ChatMessage, TranscriptionOptions, TranscriptionResult } from '../providers';
+import { App } from 'obsidian';
+import { WorkflowConfig, AIToolboxSettings } from '../settings';
 import { videoPlatformRegistry } from './video-platforms';
 import { generateFilenameTimestamp } from '../utils/date-utils';
 import {
@@ -9,119 +8,81 @@ import {
     NewNoteOutputHandler,
     AtCursorOutputHandler,
     PopupOutputHandler,
-    InputHandler,
-    InputContext,
-    InputResult,
-    VaultFileInputHandler,
-    ClipboardUrlInputHandler,
-    SelectionUrlInputHandler
+    InputResult
 } from '../handlers';
 import {
     WorkflowExecutionResult,
     WorkflowResultsMap,
     detectCircularDependency,
     hasWorkflowDependencies,
-    replaceWorkflowTokens,
-    createChatWorkflowTokens,
-    createTranscriptionWorkflowTokens,
-    getDependencyWorkflowIds,
-    gatherContextValues,
-    replaceContextTokens,
-    hasContextTokens,
-    formatTranscriptionWithTimestamps
+    getDependencyWorkflowIds
 } from './workflow-chaining';
-import { logInfo, logDebug, logWarn, logNotice, LogCategory } from '../logging';
+import {
+    ActionExecutionResult,
+    ActionResultsMap,
+    ActionExecutionContext,
+    executeAction
+} from './action-executor';
+import { logInfo, logNotice, LogCategory } from '../logging';
 
 /**
  * Create an output handler based on the workflow's output type.
  */
 function createOutputHandler(outputType: string): OutputHandler {
-	switch (outputType) {
-		case 'new-note':
-			return new NewNoteOutputHandler();
-		case 'at-cursor':
-			return new AtCursorOutputHandler();
-		case 'popup':
-		default:
-			return new PopupOutputHandler();
-	}
-}
-
-/**
- * Create an input handler based on the transcription source type.
- */
-function createInputHandler(sourceType: TranscriptionSourceType): InputHandler {
-	switch (sourceType) {
-		case 'select-file-from-vault':
-			return new VaultFileInputHandler();
-		case 'url-from-clipboard':
-			return new ClipboardUrlInputHandler();
-		case 'url-from-selection':
-			return new SelectionUrlInputHandler();
-		default:
-			return new VaultFileInputHandler();
-	}
+    switch (outputType) {
+        case 'new-note':
+            return new NewNoteOutputHandler();
+        case 'at-cursor':
+            return new AtCursorOutputHandler();
+        case 'popup':
+        default:
+            return new PopupOutputHandler();
+    }
 }
 
 /**
- * Generate a note title for transcription output based on input source.
- * - TikTok: "TikTok by <Author> - <Timestamp>"
- * - YouTube: "<Video Title> - <Author> - <Timestamp>"
- * - Vault file/fallback: "<Workflow Name> - <Timestamp>"
+ * Generate a note title based on the final action result.
  */
-function generateTranscriptionNoteTitle(inputResult: InputResult, workflowName: string): string {
-	const timestamp = generateFilenameTimestamp();
-
-	if (!inputResult.sourceUrl) {
-		return `${workflowName} - ${timestamp}`;
-	}
-
-	const handler = videoPlatformRegistry.findHandlerForUrl(inputResult.sourceUrl);
-	const metadata = inputResult.metadata;
-
-	if (handler?.platformId === 'tiktok') {
-		const author = metadata?.uploader || 'Unknown';
-		return `TikTok by ${author} - ${timestamp}`;
-	}
-
-	if (handler?.platformId === 'youtube') {
-		const title = metadata?.title || 'Video';
-		const author = metadata?.uploader || 'Unknown';
-		return `${title} - ${author} - ${timestamp}`;
-	}
-
-	return `${workflowName} - ${timestamp}`;
+function generateNoteTitle(result: ActionExecutionResult, workflowName: string): string {
+    const timestamp = generateFilenameTimestamp();
+
+    if (result.actionType === 'transcription' && result.metadata?.inputResult) {
+        const inputResult = result.metadata.inputResult;
+        if (inputResult.sourceUrl) {
+            const handler = videoPlatformRegistry.findHandlerForUrl(inputResult.sourceUrl);
+            const metadata = inputResult.metadata;
+
+            if (handler?.platformId === 'tiktok') {
+                const author = metadata?.uploader || 'Unknown';
+                return `TikTok by ${author} - ${timestamp}`;
+            }
+
+            if (handler?.platformId === 'youtube') {
+                const title = metadata?.title || 'Video';
+                const author = metadata?.uploader || 'Unknown';
+                return `${title} - ${author} - ${timestamp}`;
+            }
+        }
+    }
+
+    return `${workflowName} - ${timestamp}`;
 }
 
 /**
- * Get the prompt text for a workflow, loading from file if needed.
+ * Get the final output text from action results.
+ * Uses the response token from the last action.
  */
-async function getPromptText(app: App, workflow: WorkflowConfig): Promise<string | null> {
-	const sourceType = workflow.promptSourceType ?? 'inline';
-
-	if (sourceType === 'from-file') {
-		const filePath = workflow.promptFilePath;
-		if (!filePath || !filePath.trim()) {
-			logNotice(LogCategory.WORKFLOW, `Workflow "${workflow.name}" has no prompt file configured. Please select a file in settings.`);
-			return null;
-		}
-
-		const file = app.vault.getAbstractFileByPath(filePath);
-		if (!file || !(file instanceof TFile)) {
-			logNotice(LogCategory.WORKFLOW, `Prompt file "${filePath}" not found in vault for workflow "${workflow.name}".`);
-			return null;
-		}
-
-		try {
-			return await app.vault.read(file);
-		} catch (error) {
-			logNotice(LogCategory.WORKFLOW, `Failed to read prompt file "${filePath}": ${error instanceof Error ? error.message : String(error)}`, error);
-			return null;
-		}
-	}
-
-	// Default: use inline prompt text
-	return workflow.promptText;
+function getFinalOutputText(lastResult: ActionExecutionResult): string {
+    if (lastResult.actionType === 'chat') {
+        return lastResult.tokens['response'] || '';
+    } else {
+        // For transcription, prefer timestamped version if available
+        const timestamped = lastResult.tokens['transcriptionWithTimestamps'];
+        if (timestamped) {
+            return timestamped;
+        }
+        return lastResult.tokens['transcription'] || '';
+    }
 }
 
 /**
@@ -129,113 +90,156 @@ async function getPromptText(app: App, workflow: WorkflowConfig): Promise<string
  * Handles circular dependency detection and proper execution order.
  */
 async function executeDependencies(
-	app: App,
-	settings: AIToolboxSettings,
-	workflow: WorkflowConfig,
-	results: WorkflowResultsMap,
-	executionStack: Set<string>
+    app: App,
+    settings: AIToolboxSettings,
+    workflow: WorkflowConfig,
+    results: WorkflowResultsMap,
+    executionStack: Set<string>
 ): Promise<boolean> {
-	const dependencyIds = getDependencyWorkflowIds(workflow);
-
-	for (const depId of dependencyIds) {
-		// Skip if already executed
-		if (results.has(depId)) {
-			continue;
-		}
-
-		// Check for circular dependency (already in execution stack)
-		if (executionStack.has(depId)) {
-			const depWorkflow = settings.workflows.find(w => w.id === depId);
-			logNotice(LogCategory.WORKFLOW, `Circular dependency detected: "${depWorkflow?.name ?? depId}" is already being executed.`);
-			return false;
-		}
-
-		const depWorkflow = settings.workflows.find(w => w.id === depId);
-		if (!depWorkflow) {
-			logNotice(LogCategory.WORKFLOW, `Dependency workflow not found: ${depId}`);
-			return false;
-		}
-
-		// Add to execution stack
-		executionStack.add(depId);
-
-		// Recursively execute this dependency's dependencies first
-		if (hasWorkflowDependencies(depWorkflow)) {
-			const depSuccess = await executeDependencies(app, settings, depWorkflow, results, executionStack);
-			if (!depSuccess) {
-				return false;
-			}
-		}
-
-		// Execute the dependency workflow
-		logNotice(LogCategory.WORKFLOW, `Executing dependency: ${depWorkflow.name}...`);
-		const result = await executeWorkflowInternal(app, settings, depWorkflow, results);
-
-		if (!result.success) {
-			logNotice(LogCategory.WORKFLOW, `Dependency workflow "${depWorkflow.name}" failed: ${result.error}`);
-			return false;
-		}
-
-		results.set(depId, result);
-		executionStack.delete(depId);
-	}
-
-	return true;
+    const dependencyIds = getDependencyWorkflowIds(workflow);
+
+    for (const depId of dependencyIds) {
+        if (results.has(depId)) {
+            continue;
+        }
+
+        if (executionStack.has(depId)) {
+            const depWorkflow = settings.workflows.find(w => w.id === depId);
+            logNotice(LogCategory.WORKFLOW, `Circular dependency detected: "${depWorkflow?.name ?? depId}" is already being executed.`);
+            return false;
+        }
+
+        const depWorkflow = settings.workflows.find(w => w.id === depId);
+        if (!depWorkflow) {
+            logNotice(LogCategory.WORKFLOW, `Dependency workflow not found: ${depId}`);
+            return false;
+        }
+
+        executionStack.add(depId);
+
+        if (hasWorkflowDependencies(depWorkflow)) {
+            const depSuccess = await executeDependencies(app, settings, depWorkflow, results, executionStack);
+            if (!depSuccess) {
+                return false;
+            }
+        }
+
+        logNotice(LogCategory.WORKFLOW, `Executing dependency: ${depWorkflow.name}...`);
+        const result = await executeWorkflowInternal(app, settings, depWorkflow, results);
+
+        if (!result.success) {
+            logNotice(LogCategory.WORKFLOW, `Dependency workflow "${depWorkflow.name}" failed: ${result.error}`);
+            return false;
+        }
+
+        results.set(depId, result);
+        executionStack.delete(depId);
+    }
+
+    return true;
 }
 
 /**
- * Execute a workflow using its configured provider and display the result.
- * Routes to the appropriate execution function based on workflow type.
- * Handles workflow dependency chains by executing dependencies first.
+ * Execute a workflow by running all its actions sequentially.
+ * Each action receives the accumulated results from previous actions.
  *
  * @param app - Obsidian App instance
  * @param settings - Plugin settings
  * @param workflow - The workflow configuration to execute
  */
 export async function executeWorkflow(
-	app: App,
-	settings: AIToolboxSettings,
-	workflow: WorkflowConfig
+    app: App,
+    settings: AIToolboxSettings,
+    workflow: WorkflowConfig
 ): Promise<void> {
-	logInfo(LogCategory.WORKFLOW, `Starting workflow: ${workflow.name}`, { type: workflow.type, id: workflow.id });
-
-	// Check for circular dependencies before starting
-	if (hasWorkflowDependencies(workflow)) {
-		const cycle = detectCircularDependency(workflow, settings);
-		if (cycle.length > 0) {
-			logNotice(LogCategory.WORKFLOW, `Circular dependency detected: ${cycle.join(' → ')}`);
-			return;
-		}
-
-		// Execute all dependencies first
-		const results: WorkflowResultsMap = new Map();
-		const executionStack = new Set<string>();
-		executionStack.add(workflow.id);
-
-		const dependenciesSuccess = await executeDependencies(
-			app, settings, workflow, results, executionStack
-		);
-
-		if (!dependenciesSuccess) {
-			return;
-		}
-
-		// Execute main workflow with dependency results
-		const workflowType = workflow.type || 'chat';
-		if (workflowType === 'transcription') {
-			await executeTranscriptionWorkflow(app, settings, workflow, results);
-		} else {
-			await executeChatWorkflow(app, settings, workflow, results);
-		}
-	} else {
-		// No dependencies - execute directly
-		const workflowType = workflow.type || 'chat';
-		if (workflowType === 'transcription') {
-			await executeTranscriptionWorkflow(app, settings, workflow);
-		} else {
-			await executeChatWorkflow(app, settings, workflow);
-		}
-	}
+    logInfo(LogCategory.WORKFLOW, `Starting workflow: ${workflow.name}`, { id: workflow.id, actionCount: workflow.actions.length });
+
+    if (workflow.actions.length === 0) {
+        logNotice(LogCategory.WORKFLOW, `Workflow "${workflow.name}" has no actions. Please add actions in settings.`);
+        return;
+    }
+
+    // Check for circular dependencies before starting
+    if (hasWorkflowDependencies(workflow)) {
+        const cycle = detectCircularDependency(workflow, settings);
+        if (cycle.length > 0) {
+            logNotice(LogCategory.WORKFLOW, `Circular dependency detected: ${cycle.join(' → ')}`);
+            return;
+        }
+    }
+
+    // Execute all dependencies first
+    const dependencyResults: WorkflowResultsMap = new Map();
+    if (hasWorkflowDependencies(workflow)) {
+        const executionStack = new Set<string>();
+        executionStack.add(workflow.id);
+
+        const dependenciesSuccess = await executeDependencies(
+            app, settings, workflow, dependencyResults, executionStack
+        );
+
+        if (!dependenciesSuccess) {
+            return;
+        }
+    }
+
+    // Convert dependency results to ActionExecutionResult format
+    const dependencyActionResults: ActionResultsMap = new Map();
+    for (const [workflowId, result] of dependencyResults) {
+        dependencyActionResults.set(workflowId, {
+            actionId: workflowId,
+            actionType: result.workflowType,
+            success: result.success,
+            error: result.error,
+            tokens: result.tokens
+        });
+    }
+
+    // Execute actions sequentially
+    const actionResults: ActionResultsMap = new Map();
+    let lastResult: ActionExecutionResult | null = null;
+
+    logNotice(LogCategory.WORKFLOW, `Executing workflow: ${workflow.name}...`);
+
+    for (const action of workflow.actions) {
+        const context: ActionExecutionContext = {
+            app,
+            settings,
+            previousResults: actionResults,
+            dependencyResults: dependencyActionResults,
+            workflowName: workflow.name
+        };
+
+        const result = await executeAction(action, context);
+
+        if (!result.success) {
+            logNotice(LogCategory.WORKFLOW, `Action "${action.name}" failed: ${result.error}`);
+            return;
+        }
+
+        actionResults.set(action.id, result);
+        lastResult = result;
+    }
+
+    if (!lastResult) {
+        return;
+    }
+
+    logInfo(LogCategory.WORKFLOW, `Workflow completed: ${workflow.name}`);
+
+    // Handle output
+    const outputText = getFinalOutputText(lastResult);
+    const noteTitle = generateNoteTitle(lastResult, workflow.name);
+
+    const outputType = workflow.outputType || 'popup';
+    const handler = createOutputHandler(outputType);
+    const context: OutputContext = {
+        app,
+        workflow,
+        noteTitle
+    };
+
+    await handler.handleOutput(outputText, context);
 }
 
 /**
@@ -243,288 +247,58 @@ export async function executeWorkflow(
  * Used when executing workflows as dependencies.
  */
 async function executeWorkflowInternal(
-	app: App,
-	settings: AIToolboxSettings,
-	workflow: WorkflowConfig,
-	dependencyResults: WorkflowResultsMap
-): Promise<WorkflowExecutionResult> {
-	const workflowType = workflow.type || 'chat';
-
-	if (workflowType === 'transcription') {
-		return await executeTranscriptionWorkflowInternal(app, settings, workflow, dependencyResults);
-	} else {
-		return await executeChatWorkflowInternal(app, settings, workflow, dependencyResults);
-	}
-}
-
-/**
- * Execute a chat workflow (displays output to user).
- */
-async function executeChatWorkflow(
-	app: App,
-	settings: AIToolboxSettings,
-	workflow: WorkflowConfig,
-	dependencyResults?: WorkflowResultsMap
-): Promise<void> {
-	// Validate workflow has a provider configured
-	if (!workflow.provider) {
-		logNotice(LogCategory.WORKFLOW, `Workflow "${workflow.name}" has no provider configured. Please configure a provider in settings.`);
-		return;
-	}
-
-	// Get the prompt text (from file or inline)
-	let promptText = await getPromptText(app, workflow);
-	if (promptText === null) {
-		return; // Error already shown to user
-	}
-
-	// Replace workflow context tokens if we have dependency results
-	if (dependencyResults && dependencyResults.size > 0) {
-		promptText = replaceWorkflowTokens(promptText, dependencyResults);
-	}
-
-	// Replace context tokens ({{selection}}, {{clipboard}}, etc.) with actual values
-	if (hasContextTokens(promptText)) {
-		const contextValues = await gatherContextValues(app);
-		promptText = replaceContextTokens(promptText, contextValues);
-	}
-
-	// Validate workflow has text
-	if (!promptText.trim()) {
-		logNotice(LogCategory.WORKFLOW, `Workflow "${workflow.name}" has no prompt text. Please add prompt text in settings.`);
-		return;
-	}
-
-	// Create the provider
-	const provider = createWorkflowProvider(settings, workflow);
-	if (!provider) {
-		logNotice(LogCategory.WORKFLOW, `Could not find the configured provider for workflow "${workflow.name}". Please check your settings.`);
-		return;
-	}
-
-	// Validate provider supports chat
-	if (!provider.supportsChat()) {
-		logNotice(LogCategory.WORKFLOW, `The provider for workflow "${workflow.name}" does not support chat. Please select a chat-capable model.`);
-		return;
-	}
-
-	try {
-		logNotice(LogCategory.WORKFLOW, `Executing workflow: ${workflow.name}...`);
-		logDebug(LogCategory.WORKFLOW, `Sending chat request to provider`);
-
-		const messages: ChatMessage[] = [
-			{ role: 'user', content: promptText }
-		];
-
-		const result = await provider.sendChat(messages);
-		logInfo(LogCategory.WORKFLOW, `Chat workflow completed: ${workflow.name}`);
-
-		const outputType = workflow.outputType || 'popup';
-		const handler = createOutputHandler(outputType);
-		const context: OutputContext = {
-			app,
-			workflow,
-			promptText
-		};
-
-		await handler.handleOutput(result.content, context);
-
-	} catch (error) {
-		logNotice(LogCategory.WORKFLOW, `Workflow execution error: ${workflow.name}`, error);
-	}
-}
-
-/**
- * Internal chat workflow execution that returns a result for chaining.
- */
-async function executeChatWorkflowInternal(
-	app: App,
-	settings: AIToolboxSettings,
-	workflow: WorkflowConfig,
-	dependencyResults: WorkflowResultsMap
-): Promise<WorkflowExecutionResult> {
-	const baseResult: WorkflowExecutionResult = {
-		workflowId: workflow.id,
-		workflowType: 'chat',
-		success: false,
-		tokens: {}
-	};
-
-	if (!workflow.provider) {
-		return { ...baseResult, error: 'No provider configured' };
-	}
-
-	let promptText = await getPromptText(app, workflow);
-	if (promptText === null) {
-		return { ...baseResult, error: 'Failed to load prompt text' };
-	}
-
-	// Replace workflow context tokens
-	if (dependencyResults.size > 0) {
-		promptText = replaceWorkflowTokens(promptText, dependencyResults);
-	}
-
-	// Replace context tokens ({{selection}}, {{clipboard}}, etc.) with actual values
-	if (hasContextTokens(promptText)) {
-		const contextValues = await gatherContextValues(app);
-		promptText = replaceContextTokens(promptText, contextValues);
-	}
-
-	if (!promptText.trim()) {
-		return { ...baseResult, error: 'Empty prompt text' };
-	}
-
-	const provider = createWorkflowProvider(settings, workflow);
-	if (!provider) {
-		return { ...baseResult, error: 'Provider not found' };
-	}
-
-	if (!provider.supportsChat()) {
-		return { ...baseResult, error: 'Provider does not support chat' };
-	}
-
-	try {
-		const messages: ChatMessage[] = [
-			{ role: 'user', content: promptText }
-		];
-
-		const result = await provider.sendChat(messages);
-
-		return {
-			...baseResult,
-			success: true,
-			tokens: createChatWorkflowTokens(promptText, result.content)
-		};
-	} catch (error) {
-		const errorMessage = error instanceof Error ? error.message : String(error);
-		return { ...baseResult, error: errorMessage };
-	}
-}
-
-/**
- * Execute a transcription workflow (displays output to user).
- * Routes to the appropriate input handler based on workflow.transcriptionContext.sourceType.
- */
-async function executeTranscriptionWorkflow(
-	app: App,
-	settings: AIToolboxSettings,
-	workflow: WorkflowConfig,
-	_dependencyResults?: WorkflowResultsMap
-): Promise<void> {
-	if (!workflow.provider) {
-		logNotice(LogCategory.TRANSCRIPTION, `Workflow "${workflow.name}" has no provider configured. Please configure a provider in settings.`);
-		return;
-	}
-
-	const provider = createWorkflowProvider(settings, workflow);
-	if (!provider) {
-		logNotice(LogCategory.TRANSCRIPTION, `Could not find the configured provider for workflow "${workflow.name}". Please check your settings.`);
-		return;
-	}
-
-	if (!provider.supportsTranscription()) {
-		logNotice(LogCategory.TRANSCRIPTION, `The provider for workflow "${workflow.name}" does not support transcription. Please select a transcription-capable model.`);
-		return;
-	}
-
-	const sourceType: TranscriptionSourceType = workflow.transcriptionContext?.sourceType ?? 'url-from-clipboard';
-	const inputHandler = createInputHandler(sourceType);
-	const inputContext: InputContext = { app, settings, workflow };
-
-	try {
-		const inputResult = await inputHandler.getInput(inputContext);
-		if (!inputResult) {
-			return;
-		}
-
-		logNotice(LogCategory.TRANSCRIPTION, `Transcribing audio...`);
-		logDebug(LogCategory.TRANSCRIPTION, `Transcribing audio file: ${inputResult.audioFilePath}`);
-
-		const transcriptionOptions: TranscriptionOptions = {
-			timestampGranularity: workflow.timestampGranularity ?? 'disabled',
-			language: workflow.language || undefined
-		};
-
-		const transcriptionResult = await provider.transcribeAudio(inputResult.audioFilePath, transcriptionOptions);
-		logInfo(LogCategory.TRANSCRIPTION, `Transcription completed: ${workflow.name}`);
-
-		// Use timestamped version for display output if available, otherwise plain text
-		const formattedText = transcriptionResult.chunks.length > 0
-			? formatTranscriptionWithTimestamps(transcriptionResult.chunks)
-			: transcriptionResult.text;
-		const noteTitle = generateTranscriptionNoteTitle(inputResult, workflow.name);
-
-		const outputType = workflow.outputType || 'new-note';
-		const handler = createOutputHandler(outputType);
-		const context: OutputContext = { app, workflow, noteTitle };
-
-		await handler.handleOutput(formattedText, context);
-
-	} catch (error) {
-		logNotice(LogCategory.TRANSCRIPTION, `Transcription workflow error: ${workflow.name}`, error);
-	}
-}
-
-/**
- * Internal transcription workflow execution that returns a result for chaining.
- */
-async function executeTranscriptionWorkflowInternal(
-	app: App,
-	settings: AIToolboxSettings,
-	workflow: WorkflowConfig,
-	_dependencyResults: WorkflowResultsMap
+    app: App,
+    settings: AIToolboxSettings,
+    workflow: WorkflowConfig,
+    _dependencyResults: WorkflowResultsMap
 ): Promise<WorkflowExecutionResult> {
-	const baseResult: WorkflowExecutionResult = {
-		workflowId: workflow.id,
-		workflowType: 'transcription',
-		success: false,
-		tokens: {}
-	};
-
-	if (!workflow.provider) {
-		return { ...baseResult, error: 'No provider configured' };
-	}
-
-	const provider = createWorkflowProvider(settings, workflow);
-	if (!provider) {
-		return { ...baseResult, error: 'Provider not found' };
-	}
-
-	if (!provider.supportsTranscription()) {
-		return { ...baseResult, error: 'Provider does not support transcription' };
-	}
-
-	const sourceType: TranscriptionSourceType = workflow.transcriptionContext?.sourceType ?? 'url-from-clipboard';
-	const inputHandler = createInputHandler(sourceType);
-	const inputContext: InputContext = { app, settings, workflow };
-
-	try {
-		const inputResult = await inputHandler.getInput(inputContext);
-		if (!inputResult) {
-			return { ...baseResult, error: 'No input provided or cancelled' };
-		}
-
-		const timestampGranularity = workflow.timestampGranularity ?? 'disabled';
-		const transcriptionOptions: TranscriptionOptions = {
-			timestampGranularity,
-			language: workflow.language || undefined
-		};
-
-		const transcriptionResult = await provider.transcribeAudio(inputResult.audioFilePath, transcriptionOptions);
-
-		// Combine metadata with sourceUrl for token creation
-		const tokenMetadata = {
-			...inputResult.metadata,
-			sourceUrl: inputResult.sourceUrl
-		};
-
-		return {
-			...baseResult,
-			success: true,
-			tokens: createTranscriptionWorkflowTokens(transcriptionResult, tokenMetadata, timestampGranularity)
-		};
-	} catch (error) {
-		const errorMessage = error instanceof Error ? error.message : String(error);
-		return { ...baseResult, error: errorMessage };
-	}
+    const baseResult: WorkflowExecutionResult = {
+        workflowId: workflow.id,
+        workflowType: 'chat',
+        success: false,
+        tokens: {}
+    };
+
+    if (workflow.actions.length === 0) {
+        return { ...baseResult, error: 'No actions configured' };
+    }
+
+    // Execute actions sequentially
+    const actionResults: ActionResultsMap = new Map();
+
+    for (const action of workflow.actions) {
+        const context: ActionExecutionContext = {
+            app,
+            settings,
+            previousResults: actionResults,
+            dependencyResults: new Map(),
+            workflowName: workflow.name
+        };
+
+        const result = await executeAction(action, context);
+
+        if (!result.success) {
+            return { ...baseResult, error: result.error };
+        }
+
+        actionResults.set(action.id, result);
+    }
+
+    // Get the last action's result
+    const lastAction = workflow.actions[workflow.actions.length - 1];
+    if (!lastAction) {
+        return { ...baseResult, error: 'No actions configured' };
+    }
+    const lastResult = actionResults.get(lastAction.id);
+
+    if (!lastResult) {
+        return { ...baseResult, error: 'No result from last action' };
+    }
+
+    return {
+        workflowId: workflow.id,
+        workflowType: lastResult.actionType,
+        success: true,
+        tokens: lastResult.tokens
+    };
 }
diff --git a/src/providers/index.ts b/src/providers/index.ts
index 7cc4211..8ac64a6 100644
--- a/src/providers/index.ts
+++ b/src/providers/index.ts
@@ -23,6 +23,7 @@ export { OpenAIModelProvider } from './openai-provider';
 export {
 	createModelProvider,
 	createWorkflowProvider,
+	createActionProvider,
 	ProviderCreationError,
 } from './provider-factory';
 
diff --git a/src/providers/provider-factory.ts b/src/providers/provider-factory.ts
index cbef66b..7d294be 100644
--- a/src/providers/provider-factory.ts
+++ b/src/providers/provider-factory.ts
@@ -1,4 +1,4 @@
-import { AIToolboxSettings, AIProviderConfig, AIModelConfig, DEFAULT_OPENAI_ENDPOINT, WorkflowConfig, ProviderModelSelection } from '../settings/index';
+import { AIToolboxSettings, AIProviderConfig, AIModelConfig, DEFAULT_OPENAI_ENDPOINT, WorkflowConfig, ProviderModelSelection, WorkflowAction } from '../settings/index';
 import { ModelProvider, ModelProviderConfig } from './types';
 import { AzureOpenAIModelProvider } from './azure-openai-provider';
 import { OpenAIModelProvider } from './openai-provider';
@@ -86,9 +86,26 @@ function createProviderFromSelection(settings: AIToolboxSettings, selection: Pro
  * @throws ProviderCreationError if the provider cannot be created
  */
 export function createWorkflowProvider(settings: AIToolboxSettings, workflow: WorkflowConfig): ModelProvider | null {
-	if (!workflow.provider) {
+	// For workflows with actions, get provider from the first action
+	const firstAction = workflow.actions[0];
+	if (firstAction?.provider) {
+		return createProviderFromSelection(settings, firstAction.provider);
+	}
+	return null;
+}
+
+/**
+ * Create a provider for a specific action configuration.
+ *
+ * @param settings - Plugin settings containing provider configuration
+ * @param action - The action configuration to create a provider for
+ * @returns A configured ModelProvider for the action, or null if not configured
+ * @throws ProviderCreationError if the provider cannot be created
+ */
+export function createActionProvider(settings: AIToolboxSettings, action: WorkflowAction): ModelProvider | null {
+	if (!action.provider) {
 		return null;
 	}
 
-	return createProviderFromSelection(settings, workflow.provider);
+	return createProviderFromSelection(settings, action.provider);
 }
diff --git a/src/settings/index.ts b/src/settings/index.ts
index 77e48a2..d388a89 100644
--- a/src/settings/index.ts
+++ b/src/settings/index.ts
@@ -5,8 +5,15 @@ import { displayProvidersSettings, ProviderSettingsCallbacks } from "./providers
 import { displayWorkflowsSettings, WorkflowSettingsCallbacks } from "./workflows";
 import { displayAdditionalSettings, AdditionalSettingsCallbacks } from "./additional-settings";
 
-// Re-export all types and constants from types.ts for backward compatibility
-export { DEFAULT_OPENAI_ENDPOINT, generateId, DEFAULT_SETTINGS, DEFAULT_WORKFLOW_CONFIG } from "./types";
+// Re-export all types and constants from types.ts
+export {
+	DEFAULT_OPENAI_ENDPOINT,
+	generateId,
+	DEFAULT_SETTINGS,
+	DEFAULT_WORKFLOW_CONFIG,
+	DEFAULT_CHAT_ACTION,
+	DEFAULT_TRANSCRIPTION_ACTION
+} from "./types";
 export type {
 	AIProviderType,
 	AIModelConfig,
@@ -22,6 +29,11 @@ export type {
 	TimestampGranularity,
 	ChatContextType,
 	ChatContextConfig,
+	ActionType,
+	BaseAction,
+	ChatAction,
+	TranscriptionAction,
+	WorkflowAction,
 	AIToolboxSettings,
 	SettingsTabType,
 	ExpandOnNextRenderState
diff --git a/src/settings/types.ts b/src/settings/types.ts
index 51670bd..f5807c9 100644
--- a/src/settings/types.ts
+++ b/src/settings/types.ts
@@ -96,60 +96,111 @@ export interface ChatContextConfig {
 	type: ChatContextType;
 }
 
+
+
 /**
- * Configuration for a workflow used as a context source
+ * Action type options - determines what kind of operation an action performs
  */
-export interface WorkflowContextConfig {
-	/** The ID of the workflow to use as context */
-	workflowId: string;
-}
-// TODO: Implement circular dependency detection to prevent workflows from referencing each other in a cycle
+export type ActionType = 'chat' | 'transcription';
 
 /**
- * Configuration for a custom workflow
+ * Base action interface with common properties for all action types
  */
-export interface WorkflowConfig {
+export interface BaseAction {
+	/** Unique identifier for this action */
 	id: string;
+	/** The type of action */
+	type: ActionType;
+	/** Display name for this action */
 	name: string;
-	type: WorkflowType;
+	/** Provider and model to use for this action */
+	provider: ProviderModelSelection | null;
+}
+
+/**
+ * Chat action configuration - sends a prompt to an AI model
+ */
+export interface ChatAction extends BaseAction {
+	type: 'chat';
+	/** The prompt text to send */
 	promptText: string;
+	/** Where the prompt comes from */
 	promptSourceType: PromptSourceType;
+	/** Path to prompt file (when promptSourceType is 'from-file') */
 	promptFilePath: string;
-	provider: ProviderModelSelection | null;
-	outputType: WorkflowOutputType;
-	outputFolder: string;
-	showInCommand: boolean;
-	availableAsInput: boolean;
-	// Chat workflow context sources (optional for backward compatibility)
+	/** Context sources for this chat action */
 	contexts?: ChatContextConfig[];
-	// Workflow context sources - other workflows whose output can be used as input
-	workflowContexts?: WorkflowContextConfig[];
-	// Transcription-specific settings (optional for backward compatibility)
+}
+
+/**
+ * Transcription action configuration - transcribes audio/video
+ */
+export interface TranscriptionAction extends BaseAction {
+	type: 'transcription';
+	/** Transcription context (media type and source) */
+	transcriptionContext?: TranscriptionContextConfig;
+	/** Language for transcription */
 	language?: string;
+	/** Timestamp granularity setting */
 	timestampGranularity?: TimestampGranularity;
-	transcriptionContext?: TranscriptionContextConfig;
 }
 
 /**
- * Default configuration for a new workflow
+ * Union type for all action types
  */
-export const DEFAULT_WORKFLOW_CONFIG: Omit<WorkflowConfig, 'id'> = {
-	name: 'New workflow',
+export type WorkflowAction = ChatAction | TranscriptionAction;
+
+/**
+ * Default configuration for a new chat action
+ */
+export const DEFAULT_CHAT_ACTION: Omit<ChatAction, 'id'> = {
 	type: 'chat',
+	name: 'Chat',
+	provider: null,
 	promptText: '',
 	promptSourceType: 'inline',
 	promptFilePath: '',
+	contexts: []
+};
+
+/**
+ * Default configuration for a new transcription action
+ */
+export const DEFAULT_TRANSCRIPTION_ACTION: Omit<TranscriptionAction, 'id'> = {
+	type: 'transcription',
+	name: 'Transcription',
 	provider: null,
-	outputType: 'popup',
-	outputFolder: '',
-	showInCommand: true,
-	availableAsInput: false,
-	contexts: [],
-	workflowContexts: [],
+	transcriptionContext: { mediaType: 'video', sourceType: 'url-from-clipboard' },
 	language: '',
 	timestampGranularity: 'disabled'
 };
 
+/**
+ * Configuration for a custom workflow - a container for sequential actions
+ */
+export interface WorkflowConfig {
+	/** Unique identifier for this workflow */
+	id: string;
+	/** Display name for this workflow */
+	name: string;
+	/** Sequential list of actions to execute */
+	actions: WorkflowAction[];
+	/** How to display the final output */
+	outputType: WorkflowOutputType;
+	/** Folder for output (when outputType is 'new-note') */
+	outputFolder: string;
+}
+
+/**
+ * Default configuration for a new workflow
+ */
+export const DEFAULT_WORKFLOW_CONFIG: Omit<WorkflowConfig, 'id'> = {
+	name: 'New workflow',
+	actions: [],
+	outputType: 'popup',
+	outputFolder: ''
+};
+
 export interface AIToolboxSettings {
 	impersonateBrowser: string;
 	ytdlpLocation: string;
diff --git a/src/settings/workflows.ts b/src/settings/workflows.ts
index 6fc8b94..ce49730 100644
--- a/src/settings/workflows.ts
+++ b/src/settings/workflows.ts
@@ -1,9 +1,12 @@
-import { Notice, setIcon, Setting } from "obsidian";
+import { Setting } from "obsidian";
 import AIToolboxPlugin from "../main";
 import {
 	WorkflowConfig,
 	WorkflowOutputType,
-	WorkflowType,
+	WorkflowAction,
+	ChatAction,
+	TranscriptionAction,
+	ActionType,
 	PromptSourceType,
 	TranscriptionMediaType,
 	TranscriptionSourceType,
@@ -11,16 +14,15 @@ import {
 	ChatContextType,
 	ExpandOnNextRenderState,
 	generateId,
-	DEFAULT_WORKFLOW_CONFIG
+	DEFAULT_WORKFLOW_CONFIG,
+	DEFAULT_CHAT_ACTION,
+	DEFAULT_TRANSCRIPTION_ACTION
 } from "./types";
 import { createCollapsibleSection } from "../components/collapsible-section";
 import { createPathPicker } from "../components/path-picker";
-import { WorkflowTypeModal } from "../components/workflow-type-modal";
-import { TokenDefinition, getWorkflowContextTokens, generateWorkflowTokenTemplate } from "../tokens";
 import {
 	CHAT_CONTEXT_TYPE_LABELS,
 	CHAT_CONTEXT_TYPE_DESCRIPTIONS,
-	createContextHandler,
 	getAvailableContextTypes
 } from "../handlers";
 
@@ -62,24 +64,19 @@ export function displayWorkflowsSettings(
 	// Add workflow button
 	new Setting(containerEl)
 		.setName('Workflows')
-		.setDesc('Configure custom workflows to use with your AI providers')
+		.setDesc('Configure custom workflows with actions')
 		.addButton(button => button
 			.setButtonText('Add workflow')
 			.setCta()
-			.onClick(() => {
-				// Show modal to select workflow type
-				const modal = new WorkflowTypeModal(plugin.app, (type: WorkflowType) => {
-					const newWorkflow: WorkflowConfig = {
-						id: generateId(),
-						...DEFAULT_WORKFLOW_CONFIG,
-						type
-					};
-					plugin.settings.workflows.push(newWorkflow);
-					// Expand the newly created workflow
-					callbacks.setExpandState({ workflowId: newWorkflow.id });
-					void plugin.saveSettings().then(() => callbacks.refresh());
-				});
-				modal.open();
+			.onClick(async () => {
+				const newWorkflow: WorkflowConfig = {
+					id: generateId(),
+					...DEFAULT_WORKFLOW_CONFIG
+				};
+				plugin.settings.workflows.push(newWorkflow);
+				callbacks.setExpandState({ workflowId: newWorkflow.id });
+				await plugin.saveSettings();
+				callbacks.refresh();
 			}));
 
 	// Display each workflow
@@ -97,9 +94,14 @@ function displayWorkflowSettings(
 	const expandState = callbacks.getExpandState();
 	const shouldExpand = expandState.workflowId === workflow.id;
 
-	// Determine icon based on workflow type (default to 'chat' for backward compatibility)
-	const workflowType = workflow.type || 'chat';
-	const icon = workflowType === 'chat' ? 'message-circle' : 'audio-lines';
+	// Ensure actions array exists (for backward compatibility with old workflows)
+	if (!workflow.actions) {
+		workflow.actions = [];
+	}
+
+	// Determine icon based on first action type (or default)
+	const firstAction = workflow.actions[0];
+	const icon = firstAction?.type === 'transcription' ? 'audio-lines' : 'message-circle';
 
 	const showAdvanced = callbacks.isAdvancedVisible();
 
@@ -123,7 +125,7 @@ function displayWorkflowSettings(
 		},
 	});
 
-	// Workflow name (common to all workflow types)
+	// Workflow name
 	new Setting(contentContainer)
 		.setName('Name')
 		.setDesc('Display name for this workflow')
@@ -135,12 +137,39 @@ function displayWorkflowSettings(
 				await plugin.saveSettings();
 			}));
 
-	// Route to type-specific settings
-	if (workflowType === 'transcription') {
-		displayTranscriptionWorkflowSettings(contentContainer, plugin, workflow, callbacks, isExpanded);
-	} else {
-		// Default to chat workflow settings for backward compatibility
-		displayChatWorkflowSettings(contentContainer, plugin, workflow, callbacks, isExpanded);
+	// Actions section
+	displayActionsSection(contentContainer, plugin, workflow, callbacks, isExpanded);
+
+	// Output type
+	new Setting(contentContainer)
+		.setName('Output type')
+		.setDesc('How to display the workflow result')
+		.addDropdown(dropdown => dropdown
+			.addOptions(OUTPUT_TYPE_OPTIONS)
+			.setValue(workflow.outputType)
+			.onChange(async (value) => {
+				workflow.outputType = value as WorkflowOutputType;
+				await plugin.saveSettings();
+				if (isExpanded()) {
+					callbacks.setExpandState({ workflowId: workflow.id });
+				}
+				callbacks.refresh();
+			}));
+
+	// Output folder (only for new-note output type)
+	if (workflow.outputType === 'new-note') {
+		createPathPicker({
+			containerEl: contentContainer,
+			app: plugin.app,
+			name: 'Output folder',
+			description: 'Folder where new notes will be created',
+			initialPath: workflow.outputFolder ?? '',
+			allowFiles: false,
+			onChange: async (path: string) => {
+				workflow.outputFolder = path;
+				await plugin.saveSettings();
+			}
+		});
 	}
 
 	// Clear the expand state after rendering this workflow
@@ -150,33 +179,149 @@ function displayWorkflowSettings(
 }
 
 /**
- * Display chat-specific workflow settings
+ * Action type display labels
+ */
+const ACTION_TYPE_OPTIONS: Record<ActionType, string> = {
+	'chat': 'Chat',
+	'transcription': 'Transcription'
+};
+
+/**
+ * Display the actions section for a workflow
+ */
+function displayActionsSection(
+	containerEl: HTMLElement,
+	plugin: AIToolboxPlugin,
+	workflow: WorkflowConfig,
+	callbacks: WorkflowSettingsCallbacks,
+	isExpanded: () => boolean
+): void {
+	const actionsContainer = containerEl.createDiv('actions-section');
+	actionsContainer.createDiv({ text: 'Actions', cls: 'actions-section-title' });
+
+	// Add action button
+	new Setting(actionsContainer)
+		.addDropdown(dropdown => dropdown
+			.addOption('', 'Add action...')
+			.addOptions(ACTION_TYPE_OPTIONS)
+			.onChange(async (value) => {
+				if (!value) return;
+
+				const actionType = value as ActionType;
+				let newAction: WorkflowAction;
+
+				if (actionType === 'chat') {
+					newAction = {
+						...DEFAULT_CHAT_ACTION,
+						id: generateId(),
+						name: `Chat ${workflow.actions.length + 1}`
+					};
+				} else {
+					newAction = {
+						...DEFAULT_TRANSCRIPTION_ACTION,
+						id: generateId(),
+						name: `Transcription ${workflow.actions.length + 1}`
+					};
+				}
+
+				workflow.actions.push(newAction);
+				await plugin.saveSettings();
+
+				if (isExpanded()) {
+					callbacks.setExpandState({ workflowId: workflow.id });
+				}
+				callbacks.refresh();
+			}));
+
+	// Display each action
+	for (let i = 0; i < workflow.actions.length; i++) {
+		const action = workflow.actions[i];
+		if (!action) continue;
+		displayActionSettings(actionsContainer, plugin, workflow, action, i, callbacks, isExpanded);
+	}
+}
+
+/**
+ * Display settings for a single action
  */
-function displayChatWorkflowSettings(
-	contentContainer: HTMLElement,
+function displayActionSettings(
+	containerEl: HTMLElement,
 	plugin: AIToolboxPlugin,
 	workflow: WorkflowConfig,
+	action: WorkflowAction,
+	index: number,
 	callbacks: WorkflowSettingsCallbacks,
 	isExpanded: () => boolean
 ): void {
-	// Provider/Model selection (only models that support chat)
-	displayWorkflowProviderSelection(contentContainer, plugin, workflow);
+	const actionIcon = action.type === 'transcription' ? 'audio-lines' : 'message-circle';
+
+	const { contentContainer, updateTitle } = createCollapsibleSection({
+		containerEl,
+		title: action.name || `Action ${index + 1}`,
+		containerClass: 'action-container',
+		contentClass: 'action-content',
+		headerClass: 'action-header',
+		startExpanded: false,
+		isHeading: false,
+		icon: actionIcon,
+		onDelete: async () => {
+			workflow.actions.splice(index, 1);
+			await plugin.saveSettings();
+			if (isExpanded()) {
+				callbacks.setExpandState({ workflowId: workflow.id });
+			}
+			callbacks.refresh();
+		},
+	});
+
+	// Action name
+	new Setting(contentContainer)
+		.setName('Name')
+		.setDesc('Display name for this action')
+		.addText(text => text
+			.setValue(action.name)
+			.onChange(async (value) => {
+				action.name = value;
+				updateTitle(value || `Action ${index + 1}`);
+				await plugin.saveSettings();
+			}));
+
+	// Route to type-specific settings
+	if (action.type === 'transcription') {
+		displayTranscriptionActionSettings(contentContainer, plugin, action as TranscriptionAction, callbacks, isExpanded, workflow);
+	} else {
+		displayChatActionSettings(contentContainer, plugin, action as ChatAction, callbacks, isExpanded, workflow);
+	}
+}
+
+/**
+ * Display chat action settings
+ */
+function displayChatActionSettings(
+	containerEl: HTMLElement,
+	plugin: AIToolboxPlugin,
+	action: ChatAction,
+	callbacks: WorkflowSettingsCallbacks,
+	isExpanded: () => boolean,
+	workflow: WorkflowConfig
+): void {
+	// Provider selection
+	displayActionProviderSelection(containerEl, plugin, action, 'chat');
 
 	// Context section
-	displayContextSection(contentContainer, plugin, workflow, callbacks, isExpanded);
+	displayActionContextSection(containerEl, plugin, action, callbacks, isExpanded, workflow);
 
 	// Prompt source type dropdown
-	const promptSourceType = workflow.promptSourceType ?? 'inline';
-	new Setting(contentContainer)
+	const promptSourceType = action.promptSourceType ?? 'inline';
+	new Setting(containerEl)
 		.setName('Prompt source')
 		.setDesc('Choose where the prompt text comes from')
 		.addDropdown(dropdown => dropdown
 			.addOptions(PROMPT_SOURCE_OPTIONS)
 			.setValue(promptSourceType)
 			.onChange(async (value) => {
-				workflow.promptSourceType = value as PromptSourceType;
+				action.promptSourceType = value as PromptSourceType;
 				await plugin.saveSettings();
-				// Preserve expand state when refreshing
 				if (isExpanded()) {
 					callbacks.setExpandState({ workflowId: workflow.id });
 				}
@@ -185,15 +330,15 @@ function displayChatWorkflowSettings(
 
 	// Prompt text textarea (only show when source is inline)
 	if (promptSourceType === 'inline') {
-		new Setting(contentContainer)
+		new Setting(containerEl)
 			.setName('Prompt text')
 			.setDesc('The prompt text to send to the AI model')
 			.addTextArea(textArea => {
 				textArea
 					.setPlaceholder('Enter your prompt text here...')
-					.setValue(workflow.promptText)
+					.setValue(action.promptText ?? '')
 					.onChange(async (value) => {
-						workflow.promptText = value;
+						action.promptText = value;
 						await plugin.saveSettings();
 					});
 				textArea.inputEl.rows = 6;
@@ -204,211 +349,56 @@ function displayChatWorkflowSettings(
 	// Prompt file picker (only show when source is from-file)
 	if (promptSourceType === 'from-file') {
 		createPathPicker({
-			containerEl: contentContainer,
+			containerEl,
 			app: plugin.app,
 			name: 'Prompt file',
 			description: 'Search for a file to use as the prompt template',
 			placeholder: 'Search for file...',
-			initialPath: workflow.promptFilePath ?? '',
+			initialPath: action.promptFilePath ?? '',
 			allowFiles: true,
 			onChange: async (path: string) => {
-				workflow.promptFilePath = path;
+				action.promptFilePath = path;
 				await plugin.saveSettings();
 			}
 		});
 	}
-
-	// Make available as input to other workflows toggle
-	new Setting(contentContainer)
-		.setName('Make available as input to other workflows')
-		.setDesc('Allow this workflow to be used as input context for other workflows')
-		.addToggle(toggle => toggle
-			.setValue(workflow.availableAsInput ?? false)
-			.onChange(async (value) => {
-				workflow.availableAsInput = value;
-				await plugin.saveSettings();
-				// Refresh so other workflow dropdowns update to include/exclude this workflow
-				if (isExpanded()) {
-					callbacks.setExpandState({ workflowId: workflow.id });
-				}
-				callbacks.refresh();
-			}));
-
-	// List in Workflow Command toggle
-	new Setting(contentContainer)
-		.setName('List in run workflow command')
-		.setDesc('Show this workflow in the list when running the run workflow command')
-		.addToggle(toggle => toggle
-			.setValue(workflow.showInCommand ?? true)
-			.onChange(async (value) => {
-				workflow.showInCommand = value;
-				await plugin.saveSettings();
-				// Preserve expand state when refreshing
-				if (isExpanded()) {
-					callbacks.setExpandState({ workflowId: workflow.id });
-				}
-				callbacks.refresh();
-			}));
-
-	// Output type selection (only show if workflow is listed in command)
-	if (workflow.showInCommand ?? true) {
-		new Setting(contentContainer)
-			.setName('Output type')
-			.setDesc('Choose how to display the AI response')
-			.addDropdown(dropdown => dropdown
-				.addOptions(OUTPUT_TYPE_OPTIONS)
-				.setValue(workflow.outputType || 'popup')
-				.onChange(async (value) => {
-					workflow.outputType = value as WorkflowOutputType;
-					await plugin.saveSettings();
-					// Preserve expand state when refreshing (output folder visibility may change)
-					if (isExpanded()) {
-						callbacks.setExpandState({ workflowId: workflow.id });
-					}
-					callbacks.refresh();
-				}));
-
-		// Output folder (only show if output type is new-note)
-		if (workflow.outputType === 'new-note') {
-			createPathPicker({
-				containerEl: contentContainer,
-				app: plugin.app,
-				name: 'Output folder',
-				description: 'Folder where notes will be created (leave empty to use default)',
-				placeholder: 'Search for folder...',
-				initialPath: workflow.outputFolder || '',
-				onChange: async (path: string) => {
-					workflow.outputFolder = path;
-					await plugin.saveSettings();
-				}
-			});
-		}
-	}
-}
-
-function displayWorkflowProviderSelection(
-	containerEl: HTMLElement,
-	plugin: AIToolboxPlugin,
-	workflow: WorkflowConfig
-): void {
-	const providers = plugin.settings.providers;
-	const currentSelection = workflow.provider;
-
-	// Build options for provider/model dropdown (only models that support chat)
-	const options: Record<string, string> = { '': 'Select a provider and model' };
-	for (const provider of providers) {
-		for (const model of provider.models) {
-			if (model.supportsChat) {
-				const key = `${provider.id}:${model.id}`;
-				options[key] = `${provider.name} - ${model.name}`;
-			}
-		}
-	}
-
-	const currentValue = currentSelection
-		? `${currentSelection.providerId}:${currentSelection.modelId}`
-		: '';
-
-	new Setting(containerEl)
-		.setName('Provider')
-		.setDesc('Select the provider and model to use for this workflow')
-		.addDropdown(dropdown => dropdown
-			.addOptions(options)
-			.setValue(currentValue)
-			.onChange(async (value) => {
-				if (value === '') {
-					workflow.provider = null;
-				} else {
-					const parts = value.split(':');
-					if (parts.length === 2 && parts[0] && parts[1]) {
-						workflow.provider = {
-							providerId: parts[0],
-							modelId: parts[1]
-						};
-					}
-				}
-				await plugin.saveSettings();
-			}));
 }
 
 /**
- * Display provider selection for transcription workflows
- * Only shows models that support transcription
+ * Display transcription action settings
  */
-function displayTranscriptionProviderSelection(
+function displayTranscriptionActionSettings(
 	containerEl: HTMLElement,
 	plugin: AIToolboxPlugin,
+	action: TranscriptionAction,
+	callbacks: WorkflowSettingsCallbacks,
+	isExpanded: () => boolean,
 	workflow: WorkflowConfig
 ): void {
-	const providers = plugin.settings.providers;
-	const currentSelection = workflow.provider;
+	// Provider selection
+	displayActionProviderSelection(containerEl, plugin, action, 'transcription');
 
-	// Build options for provider/model dropdown (only models that support transcription)
-	const options: Record<string, string> = { '': 'Select a provider and model' };
-	for (const provider of providers) {
-		for (const model of provider.models) {
-			if (model.supportsTranscription) {
-				const key = `${provider.id}:${model.id}`;
-				options[key] = `${provider.name} - ${model.name}`;
-			}
-		}
+	// Ensure transcriptionContext exists
+	if (!action.transcriptionContext) {
+		action.transcriptionContext = { mediaType: 'video', sourceType: 'url-from-clipboard' };
 	}
 
-	const currentValue = currentSelection
-		? `${currentSelection.providerId}:${currentSelection.modelId}`
-		: '';
-
-	new Setting(containerEl)
-		.setName('Provider')
-		.setDesc('Select the provider and model to use for transcription')
-		.addDropdown(dropdown => dropdown
-			.addOptions(options)
-			.setValue(currentValue)
-			.onChange(async (value) => {
-				if (value === '') {
-					workflow.provider = null;
-				} else {
-					const parts = value.split(':');
-					if (parts.length === 2 && parts[0] && parts[1]) {
-						workflow.provider = {
-							providerId: parts[0],
-							modelId: parts[1]
-						};
-					}
-				}
-				await plugin.saveSettings();
-			}));
-}
-
-/**
- * Display transcription-specific workflow settings
- */
-function displayTranscriptionWorkflowSettings(
-	contentContainer: HTMLElement,
-	plugin: AIToolboxPlugin,
-	workflow: WorkflowConfig,
-	callbacks: WorkflowSettingsCallbacks,
-	isExpanded: () => boolean
-): void {
-	// Provider selection (only models that support transcription)
-	displayTranscriptionProviderSelection(contentContainer, plugin, workflow);
-
 	// Media type dropdown
 	const mediaTypeOptions: Record<TranscriptionMediaType, string> = {
 		'video': 'Video',
 		'audio': 'Audio'
 	};
-	new Setting(contentContainer)
+	new Setting(containerEl)
 		.setName('Media type')
 		.setDesc('The type of media to transcribe')
 		.addDropdown(dropdown => dropdown
 			.addOptions(mediaTypeOptions)
-			.setValue(workflow.transcriptionContext?.mediaType ?? 'video')
+			.setValue(action.transcriptionContext?.mediaType ?? 'video')
 			.onChange(async (value) => {
-				if (!workflow.transcriptionContext) {
-					workflow.transcriptionContext = { mediaType: 'video', sourceType: 'url-from-clipboard' };
+				if (!action.transcriptionContext) {
+					action.transcriptionContext = { mediaType: 'video', sourceType: 'url-from-clipboard' };
 				}
-				workflow.transcriptionContext.mediaType = value as TranscriptionMediaType;
+				action.transcriptionContext.mediaType = value as TranscriptionMediaType;
 				await plugin.saveSettings();
 			}));
 
@@ -418,30 +408,30 @@ function displayTranscriptionWorkflowSettings(
 		'url-from-clipboard': 'URL from clipboard',
 		'url-from-selection': 'URL from selection'
 	};
-	new Setting(contentContainer)
+	new Setting(containerEl)
 		.setName('Source')
 		.setDesc('Where to get the media file from')
 		.addDropdown(dropdown => dropdown
 			.addOptions(sourceTypeOptions)
-			.setValue(workflow.transcriptionContext?.sourceType ?? 'url-from-clipboard')
+			.setValue(action.transcriptionContext?.sourceType ?? 'url-from-clipboard')
 			.onChange(async (value) => {
-				if (!workflow.transcriptionContext) {
-					workflow.transcriptionContext = { mediaType: 'video', sourceType: 'url-from-clipboard' };
+				if (!action.transcriptionContext) {
+					action.transcriptionContext = { mediaType: 'video', sourceType: 'url-from-clipboard' };
 				}
-				workflow.transcriptionContext.sourceType = value as TranscriptionSourceType;
+				action.transcriptionContext.sourceType = value as TranscriptionSourceType;
 				await plugin.saveSettings();
 			}));
 
 	// Language setting (advanced)
 	const showAdvanced = callbacks.isAdvancedVisible();
-	const languageSetting = new Setting(contentContainer)
+	const languageSetting = new Setting(containerEl)
 		.setName('Language')
 		.setDesc('Optional language code for transcription (e.g., "en", "es", "fr"). Leave empty for auto-detection.')
 		.addText(text => text
 			.setPlaceholder('Auto-detect')
-			.setValue(workflow.language || '')
+			.setValue(action.language ?? '')
 			.onChange(async (value) => {
-				workflow.language = value;
+				action.language = value;
 				await plugin.saveSettings();
 			}));
 	languageSetting.settingEl.toggleClass('settings-advanced-hidden', !showAdvanced);
@@ -455,146 +445,124 @@ function displayTranscriptionWorkflowSettings(
 		'segment': 'Segment (sentence/phrase level)',
 		'word': 'Word (individual word level)'
 	};
-	const granularitySetting = new Setting(contentContainer)
+	const granularitySetting = new Setting(containerEl)
 		.setName('Timestamp granularity')
 		.setDesc('Level of detail for timestamps. Disabling reduces token usage.')
 		.addDropdown(dropdown => dropdown
 			.addOptions(granularityOptions)
-			.setValue(workflow.timestampGranularity ?? 'disabled')
+			.setValue(action.timestampGranularity ?? 'disabled')
 			.onChange(async (value) => {
-				workflow.timestampGranularity = value as TimestampGranularity;
+				action.timestampGranularity = value as TimestampGranularity;
 				await plugin.saveSettings();
-				callbacks.setExpandState({ workflowId: workflow.id });
+				if (isExpanded()) {
+					callbacks.setExpandState({ workflowId: workflow.id });
+				}
 				callbacks.refresh();
 			}));
 	granularitySetting.settingEl.toggleClass('settings-advanced-hidden', !showAdvanced);
 	if (showAdvanced) {
 		granularitySetting.nameEl.addClass('settings-advanced-name');
 	}
+}
 
-	// Make available as input to other workflows toggle
-	new Setting(contentContainer)
-		.setName('Make available as input to other workflows')
-		.setDesc('Allow this workflow to be used as input context for other workflows')
-		.addToggle(toggle => toggle
-			.setValue(workflow.availableAsInput ?? false)
-			.onChange(async (value) => {
-				workflow.availableAsInput = value;
-				await plugin.saveSettings();
-				// Refresh so other workflow dropdowns update to include/exclude this workflow
-				if (isExpanded()) {
-					callbacks.setExpandState({ workflowId: workflow.id });
-				}
-				callbacks.refresh();
-			}));
+/**
+ * Display provider selection for an action
+ */
+function displayActionProviderSelection(
+	containerEl: HTMLElement,
+	plugin: AIToolboxPlugin,
+	action: WorkflowAction,
+	capability: 'chat' | 'transcription'
+): void {
+	const providers = plugin.settings.providers;
+	const currentSelection = action.provider;
 
-	// List in Workflow Command toggle
-	new Setting(contentContainer)
-		.setName('List in run workflow command')
-		.setDesc('Show this workflow in the list when running the run workflow command')
-		.addToggle(toggle => toggle
-			.setValue(workflow.showInCommand ?? true)
-			.onChange(async (value) => {
-				workflow.showInCommand = value;
-				await plugin.saveSettings();
-				// Preserve expand state when refreshing
-				if (isExpanded()) {
-					callbacks.setExpandState({ workflowId: workflow.id });
-				}
-				callbacks.refresh();
-			}));
+	// Build options for provider/model dropdown
+	const options: Record<string, string> = { '': 'Select a provider and model' };
+	for (const provider of providers) {
+		for (const model of provider.models) {
+			const supportsCapability = capability === 'chat' ? model.supportsChat : model.supportsTranscription;
+			if (supportsCapability) {
+				const key = `${provider.id}:${model.id}`;
+				options[key] = `${provider.name} - ${model.name}`;
+			}
+		}
+	}
 
-	// Output type selection (only show if workflow is listed in command)
-	if (workflow.showInCommand ?? true) {
-		new Setting(contentContainer)
-			.setName('Output type')
-			.setDesc('Choose how to display the transcription result')
-			.addDropdown(dropdown => dropdown
-				.addOptions(OUTPUT_TYPE_OPTIONS)
-				.setValue(workflow.outputType || 'new-note')
-				.onChange(async (value) => {
-					workflow.outputType = value as WorkflowOutputType;
-					await plugin.saveSettings();
-					// Preserve expand state when refreshing (output folder visibility may change)
-					if (isExpanded()) {
-						callbacks.setExpandState({ workflowId: workflow.id });
-					}
-					callbacks.refresh();
-				}));
+	const currentValue = currentSelection
+		? `${currentSelection.providerId}:${currentSelection.modelId}`
+		: '';
 
-		// Output folder (only show if output type is new-note)
-		if (workflow.outputType === 'new-note') {
-			createPathPicker({
-				containerEl: contentContainer,
-				app: plugin.app,
-				name: 'Output folder',
-				description: 'Folder where transcription notes will be created (leave empty to use default)',
-				placeholder: 'Search for folder...',
-				initialPath: workflow.outputFolder || '',
-				onChange: async (path: string) => {
-					workflow.outputFolder = path;
-					await plugin.saveSettings();
+	new Setting(containerEl)
+		.setName('Provider')
+		.setDesc(`Select the provider and model to use for ${capability}`)
+		.addDropdown(dropdown => dropdown
+			.addOptions(options)
+			.setValue(currentValue)
+			.onChange(async (value) => {
+				if (value === '') {
+					action.provider = null;
+				} else {
+					const parts = value.split(':');
+					if (parts.length === 2 && parts[0] && parts[1]) {
+						action.provider = {
+							providerId: parts[0],
+							modelId: parts[1]
+						};
+					}
 				}
-			});
-		}
-	}
+				await plugin.saveSettings();
+			}));
 }
 
 /**
- * Display context section for chat workflows
+ * Display context section for chat actions
  */
-function displayContextSection(
-	contentContainer: HTMLElement,
+function displayActionContextSection(
+	containerEl: HTMLElement,
 	plugin: AIToolboxPlugin,
-	workflow: WorkflowConfig,
+	action: ChatAction,
 	callbacks: WorkflowSettingsCallbacks,
-	isExpanded: () => boolean
+	isExpanded: () => boolean,
+	workflow: WorkflowConfig
 ): void {
 	// Ensure contexts array exists
-	if (!workflow.contexts) {
-		workflow.contexts = [];
+	if (!action.contexts) {
+		action.contexts = [];
 	}
 
 	// Helper to check if a context type is enabled
 	const isContextEnabled = (type: ChatContextType): boolean => {
-		return workflow.contexts?.some(c => c.type === type) ?? false;
+		return action.contexts?.some(c => c.type === type) ?? false;
 	};
 
 	// Helper to toggle a context type
 	const toggleContext = async (type: ChatContextType, enabled: boolean): Promise<void> => {
-		if (!workflow.contexts) {
-			workflow.contexts = [];
+		if (!action.contexts) {
+			action.contexts = [];
 		}
 
 		if (enabled) {
-			// Add context if not already present
-			if (!workflow.contexts.some(c => c.type === type)) {
-				workflow.contexts.push({
+			if (!action.contexts.some(c => c.type === type)) {
+				action.contexts.push({
 					id: generateId(),
 					type
 				});
 			}
 		} else {
-			// Remove context
-			workflow.contexts = workflow.contexts.filter(c => c.type !== type);
+			action.contexts = action.contexts.filter(c => c.type !== type);
 		}
 
 		await plugin.saveSettings();
 
-		// Refresh to update the available tokens section
 		if (isExpanded()) {
-			// Check if available tokens section is expanded before refreshing
-			const tokensContent = contentContainer.querySelector('.available-tokens-content');
-			const tokensExpanded = tokensContent?.classList.contains('is-expanded') ?? false;
-			callbacks.setExpandState({ workflowId: workflow.id, availableTokensExpanded: tokensExpanded });
+			callbacks.setExpandState({ workflowId: workflow.id });
 		}
 		callbacks.refresh();
 	};
 
 	// Context section container
-	const contextSection = contentContainer.createDiv('context-section');
-
-	// Section header
+	const contextSection = containerEl.createDiv('context-section');
 	contextSection.createDiv({ text: 'Context', cls: 'context-section-title' });
 
 	// Toggles row
@@ -617,273 +585,4 @@ function displayContextSection(
 					await toggleContext(contextType, value);
 				}));
 	}
-
-	// Workflow context section (select other workflows as input)
-	displayWorkflowContextSection(contextSection, plugin, workflow, callbacks, isExpanded);
-
-	// Available tokens section (inside the context section)
-	displayAvailableTokensSection(contextSection, workflow, plugin, callbacks);
-}
-
-/**
- * Display workflow context section for selecting other workflows as input
- */
-function displayWorkflowContextSection(
-	containerEl: HTMLElement,
-	plugin: AIToolboxPlugin,
-	workflow: WorkflowConfig,
-	callbacks: WorkflowSettingsCallbacks,
-	isExpanded: () => boolean
-): void {
-	// Ensure workflowContexts array exists
-	if (!workflow.workflowContexts) {
-		workflow.workflowContexts = [];
-	}
-
-	// Get workflows that are available as input (excluding current workflow)
-	const availableWorkflows = plugin.settings.workflows.filter(
-		w => w.availableAsInput && w.id !== workflow.id
-	);
-
-	// Workflow context subsection
-	const workflowSection = containerEl.createDiv('workflow-context-section');
-
-	// Dropdown to select a workflow
-	const dropdownContainer = workflowSection.createDiv('workflow-context-dropdown-container');
-
-	// Build options for dropdown
-	const selectedWorkflowIds = new Set(workflow.workflowContexts?.map(wc => wc.workflowId) ?? []);
-	const unselectedWorkflows = availableWorkflows.filter(w => !selectedWorkflowIds.has(w.id));
-
-	// Determine if dropdown should be disabled
-	const isDisabled = unselectedWorkflows.length === 0;
-
-	new Setting(dropdownContainer)
-		.setName('Add workflow')
-		.setDesc('Select a workflow to use as context')
-		.addDropdown(dropdown => {
-			dropdown.addOption('', 'Select a workflow...');
-			for (const w of unselectedWorkflows) {
-				dropdown.addOption(w.id, w.name);
-			}
-			dropdown.setDisabled(isDisabled);
-			if (isDisabled) {
-				dropdown.selectEl.setAttribute('title', 'No additional valid workflows are available.');
-			}
-			dropdown.onChange(async (value) => {
-				if (!value) return;
-
-				if (!workflow.workflowContexts) {
-					workflow.workflowContexts = [];
-				}
-				workflow.workflowContexts.push({ workflowId: value });
-				await plugin.saveSettings();
-
-				if (isExpanded()) {
-					// Check if available tokens section is expanded before refreshing
-					const tokensContent = workflowSection.parentElement?.querySelector('.available-tokens-content');
-					const tokensExpanded = tokensContent?.classList.contains('is-expanded') ?? false;
-					callbacks.setExpandState({ workflowId: workflow.id, availableTokensExpanded: tokensExpanded });
-				}
-				callbacks.refresh();
-			});
-		});
-
-	// Display selected workflows with delete buttons
-	const selectedWorkflowsContainer = workflowSection.createDiv('workflow-context-selected-list');
-
-	for (const workflowContext of workflow.workflowContexts ?? []) {
-		const selectedWorkflow = plugin.settings.workflows.find(w => w.id === workflowContext.workflowId);
-		if (!selectedWorkflow) continue;
-
-		const itemContainer = selectedWorkflowsContainer.createDiv('workflow-context-selected-item');
-
-		const itemLabel = itemContainer.createSpan('workflow-context-selected-label');
-		itemLabel.textContent = selectedWorkflow.name;
-
-		const deleteButton = itemContainer.createEl('button', { cls: 'workflow-context-delete-button' });
-		deleteButton.textContent = '×';
-		deleteButton.setAttribute('aria-label', `Remove ${selectedWorkflow.name}`);
-		deleteButton.addEventListener('click', () => {
-			workflow.workflowContexts = workflow.workflowContexts?.filter(
-				wc => wc.workflowId !== workflowContext.workflowId
-			);
-
-			void plugin.saveSettings().then(() => {
-				if (isExpanded()) {
-					// Check if available tokens section is expanded before refreshing
-					const tokensContent = workflowSection.parentElement?.querySelector('.available-tokens-content');
-					const tokensExpanded = tokensContent?.classList.contains('is-expanded') ?? false;
-					callbacks.setExpandState({ workflowId: workflow.id, availableTokensExpanded: tokensExpanded });
-				}
-				callbacks.refresh();
-			});
-		});
-	}
-}
-
-/**
- * Represents a group of tokens from a single source
- */
-interface TokenGroup {
-	name: string;
-	tokens: TokenDefinition[];
-	/** For workflow groups: the source workflow ID */
-	workflowId?: string;
-	/** For workflow groups: the source workflow type */
-	workflowType?: WorkflowType;
-	/** For transcription workflows: the timestamp granularity setting */
-	timestampGranularity?: TimestampGranularity;
-}
-
-/**
- * Render a clickable token reference element with copy functionality
- */
-function renderTokenReference(
-	container: HTMLElement,
-	token: TokenDefinition
-): void {
-	const listItem = container.createEl('li', { cls: 'available-tokens-item' });
-
-	const tokenText = `{{${token.name}}}`;
-	const tokenRef = listItem.createEl('code', { cls: 'available-tokens-reference' });
-	tokenRef.textContent = tokenText;
-	tokenRef.setAttribute('title', 'Click to copy');
-	tokenRef.addEventListener('click', (e) => {
-		e.stopPropagation();
-		void navigator.clipboard.writeText(tokenText).then(() => {
-			new Notice(`Copied ${tokenText}`);
-		});
-	});
-
-	const tokenDesc = listItem.createSpan('available-tokens-description');
-	tokenDesc.textContent = ` — ${token.description}`;
-}
-
-/**
- * Display the available tokens section showing tokens from all configured contexts
- */
-function displayAvailableTokensSection(
-	containerEl: HTMLElement,
-	workflow: WorkflowConfig,
-	plugin: AIToolboxPlugin,
-	callbacks: WorkflowSettingsCallbacks
-): void {
-	const contexts = workflow.contexts ?? [];
-	const workflowContexts = workflow.workflowContexts ?? [];
-
-	// Check if there's anything to display
-	const hasContexts = contexts.length > 0;
-	const hasWorkflowContexts = workflowContexts.length > 0;
-
-	if (!hasContexts && !hasWorkflowContexts) {
-		return;
-	}
-
-	// Collect token groups
-	const tokenGroups: TokenGroup[] = [];
-	let totalTokenCount = 0;
-
-	// Add context handler tokens as a single "Context" group
-	if (hasContexts) {
-		const contextTokens: TokenDefinition[] = [];
-		for (const context of contexts) {
-			const handler = createContextHandler(context.type);
-			const tokens = handler.getAvailableTokens();
-			contextTokens.push(...tokens);
-		}
-		if (contextTokens.length > 0) {
-			tokenGroups.push({ name: 'Context', tokens: contextTokens });
-			totalTokenCount += contextTokens.length;
-		}
-	}
-
-	// Add workflow context tokens grouped by source workflow
-	for (const workflowContext of workflowContexts) {
-		const sourceWorkflow = plugin.settings.workflows.find(w => w.id === workflowContext.workflowId);
-		if (!sourceWorkflow) continue;
-
-		const workflowType = sourceWorkflow.type || 'chat';
-		const timestampGranularity = sourceWorkflow.timestampGranularity;
-		const workflowTokens = getWorkflowContextTokens(
-			sourceWorkflow.id,
-			workflowType,
-			{ timestampGranularity }
-		);
-		if (workflowTokens.length > 0) {
-			tokenGroups.push({
-				name: sourceWorkflow.name || 'Unnamed workflow',
-				tokens: workflowTokens,
-				workflowId: sourceWorkflow.id,
-				workflowType: workflowType,
-				timestampGranularity: timestampGranularity
-			});
-			totalTokenCount += workflowTokens.length;
-		}
-	}
-
-	if (totalTokenCount === 0) {
-		return;
-	}
-
-	// Check if we should expand the tokens section
-	const expandState = callbacks.getExpandState();
-	const shouldExpand = expandState.availableTokensExpanded === true;
-
-	// Create collapsible available tokens section
-	const sectionContainer = containerEl.createDiv('available-tokens-section');
-
-	const header = sectionContainer.createDiv('available-tokens-header');
-	const headerArrow = header.createSpan('available-tokens-arrow');
-	headerArrow.textContent = shouldExpand ? '▾' : '▸';
-
-	const headerTitle = header.createSpan('available-tokens-title');
-	headerTitle.textContent = `Available tokens (${totalTokenCount})`;
-
-	const contentClasses = shouldExpand ? 'available-tokens-content is-expanded' : 'available-tokens-content is-collapsed';
-	const content = sectionContainer.createDiv(contentClasses);
-
-	const description = content.createDiv('available-tokens-description-text');
-	description.textContent = 'These tokens available for use in prompts and output templates downstream.';
-
-	// Render each token group
-	for (const group of tokenGroups) {
-		const groupContainer = content.createDiv('available-tokens-group');
-
-		const groupHeader = groupContainer.createDiv('available-tokens-group-header');
-
-		// Make workflow group headers clickable to copy full template
-		if (group.workflowId && group.workflowType) {
-			groupHeader.createSpan({ text: group.name });
-			const copyIcon = groupHeader.createSpan({ cls: 'available-tokens-copy-icon' });
-			setIcon(copyIcon, 'copy');
-			groupHeader.addClass('is-clickable');
-			groupHeader.setAttribute('title', 'Click to copy all tokens as template');
-			const workflowId = group.workflowId;
-			const workflowType = group.workflowType;
-			const timestampGranularity = group.timestampGranularity;
-			groupHeader.addEventListener('click', (e) => {
-				e.stopPropagation();
-				const template = generateWorkflowTokenTemplate(workflowId, workflowType, { timestampGranularity });
-				void navigator.clipboard.writeText(template).then(() => {
-					new Notice('Copied token template to clipboard');
-				});
-			});
-		} else {
-			groupHeader.textContent = group.name;
-		}
-
-		const tokenList = groupContainer.createEl('ul', { cls: 'available-tokens-list' });
-		for (const token of group.tokens) {
-			renderTokenReference(tokenList, token);
-		}
-	}
-
-	// Toggle handler
-	header.addEventListener('click', () => {
-		const isCollapsed = content.classList.contains('is-collapsed');
-		content.classList.toggle('is-collapsed', !isCollapsed);
-		content.classList.toggle('is-expanded', isCollapsed);
-		headerArrow.textContent = isCollapsed ? '▾' : '▸';
-	});
 }
diff --git a/src/tokens/types.ts b/src/tokens/types.ts
index a18b6a8..ffdb010 100644
--- a/src/tokens/types.ts
+++ b/src/tokens/types.ts
@@ -1,4 +1,4 @@
-import { WorkflowType, TimestampGranularity } from "../settings/types";
+import { ActionType, TimestampGranularity } from "../settings/types";
 
 /**
  * Token definition for display in the settings UI
@@ -88,21 +88,20 @@ const TOKEN_LABELS: Record<string, string> = {
  * Options for getting token definitions
  */
 export interface TokenDefinitionOptions {
-	/** For transcription workflows, the timestamp granularity setting */
+	/** For transcription actions, the timestamp granularity setting */
 	timestampGranularity?: TimestampGranularity;
 }
 
 /**
- * Get token definitions for a workflow type.
- * For transcription workflows, the transcriptionWithTimestamps token is excluded
+ * Get token definitions for an action type.
+ * For transcription actions, the transcriptionWithTimestamps token is excluded
  * when timestampGranularity is 'disabled' or undefined (default).
  */
-export function getTokenDefinitionsForType(
-	type: WorkflowType,
+export function getTokenDefinitionsForActionType(
+	type: ActionType,
 	options?: TokenDefinitionOptions
 ): TokenDefinition[] {
 	if (type === 'transcription') {
-		// Filter out transcriptionWithTimestamps when granularity is disabled or not set
 		const granularity = options?.timestampGranularity ?? 'disabled';
 		if (granularity === 'disabled') {
 			return TRANSCRIPTION_WORKFLOW_TOKENS.filter(
@@ -114,20 +113,41 @@ export function getTokenDefinitionsForType(
 	return CHAT_WORKFLOW_TOKENS;
 }
 
+/**
+ * Get token definitions for an action, prefixed with action ID.
+ * Used to show available tokens in the settings UI.
+ *
+ * @param actionId - The ID of the action
+ * @param actionType - The type of the action (chat or transcription)
+ * @param options - Optional settings like timestampGranularity
+ */
+export function getActionTokens(
+	actionId: string,
+	actionType: ActionType,
+	options?: TokenDefinitionOptions
+): TokenDefinition[] {
+	const baseTokens = getTokenDefinitionsForActionType(actionType, options);
+
+	return baseTokens.map(token => ({
+		name: `${actionId}.${token.name}`,
+		description: token.description
+	}));
+}
+
 /**
  * Get token definitions for a workflow used as a context source.
- * Returns tokens prefixed with the workflow ID to avoid collisions.
+ * Returns tokens from the last action, prefixed with the workflow ID.
  *
  * @param workflowId - The ID of the workflow
- * @param workflowType - The type of the workflow (chat or transcription)
+ * @param lastActionType - The type of the last action in the workflow
  * @param options - Optional settings like timestampGranularity
  */
 export function getWorkflowContextTokens(
 	workflowId: string,
-	workflowType: WorkflowType,
+	lastActionType: ActionType,
 	options?: TokenDefinitionOptions
 ): TokenDefinition[] {
-	const baseTokens = getTokenDefinitionsForType(workflowType, options);
+	const baseTokens = getTokenDefinitionsForActionType(lastActionType, options);
 
 	return baseTokens.map(token => ({
 		name: `${workflowId}.${token.name}`,
@@ -140,15 +160,15 @@ export function getWorkflowContextTokens(
  * Used for copying a complete set of token references to the clipboard.
  *
  * @param workflowId - The ID of the workflow
- * @param workflowType - The type of the workflow (chat or transcription)
+ * @param lastActionType - The type of the last action in the workflow
  * @param options - Optional settings like timestampGranularity
  */
 export function generateWorkflowTokenTemplate(
 	workflowId: string,
-	workflowType: WorkflowType,
+	lastActionType: ActionType,
 	options?: TokenDefinitionOptions
 ): string {
-	const tokens = getTokenDefinitionsForType(workflowType, options);
+	const tokens = getTokenDefinitionsForActionType(lastActionType, options);
 
 	return tokens
 		.map(token => {
diff --git a/styles.css b/styles.css
index d1d0cc8..b4df258 100644
--- a/styles.css
+++ b/styles.css
@@ -185,6 +185,33 @@ If your plugin does not need CSS, delete this file.
 	padding-top: 8px;
 }
 
+/* Action container styles */
+.action-container {
+	margin-left: 20px;
+	margin-bottom: 8px;
+}
+
+.action-header {
+	cursor: pointer;
+	background: var(--background-secondary);
+	border-radius: 4px;
+	padding: 4px 8px;
+}
+
+.action-header:hover .setting-item-name {
+	color: var(--text-accent);
+}
+
+.action-content.is-collapsed {
+	display: none;
+}
+
+.action-content.is-expanded {
+	display: block;
+	padding-top: 8px;
+	padding-left: 12px;
+}
+
 .workflow-textarea {
 	width: 100%;
 	min-height: 120px;

From 866712c3f43a5c9a947e4448437cacc4d3777565 Mon Sep 17 00:00:00 2001
From: dalinicus <contact@dalinicus.com>
Date: Sun, 11 Jan 2026 22:54:54 -0600
Subject: [PATCH 7/7] actions system

---
 AGENTS.md                                     | 204 +++++++++++--
 manifest.json                                 |   2 +-
 src/components/path-picker.ts                 |   2 +-
 src/handlers/index.ts                         |   3 +-
 src/handlers/input/index.ts                   |   1 +
 src/handlers/input/token-url-input-handler.ts |  37 +++
 src/logging/log-manager.ts                    |  14 +-
 src/logging/log-pane-view.ts                  |   8 +-
 src/main.ts                                   |   7 +-
 src/processing/action-executor.ts             |  80 ++++--
 src/processing/video-processor.ts             |   6 +
 src/processing/workflow-chaining.ts           |  41 ++-
 src/processing/workflow-executor.ts           |  19 +-
 src/settings/additional-settings.ts           |  19 ++
 src/settings/index.ts                         |   1 -
 src/settings/types.ts                         |  13 +-
 src/settings/workflows.ts                     | 269 ++++++++++--------
 src/tokens/types.ts                           | 113 ++++----
 styles.css                                    |  23 +-
 versions.json                                 |   3 +-
 20 files changed, 583 insertions(+), 282 deletions(-)
 create mode 100644 src/handlers/input/token-url-input-handler.ts

diff --git a/AGENTS.md b/AGENTS.md
index 6212a38..9a9f753 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -150,9 +150,16 @@ The `versions.json` file maps plugin versions to minimum Obsidian app versions.
 - Remove all other minor and patch versions for that major version
 - Example: If releasing 2.0.0, keep only 1.2.3 and 1.3.2 (remove 1.0.0, 1.1.0, 1.2.1, 1.2.2, 1.3.0, 1.3.1, etc.)
 
-#### 3. Run linting
-- Run `eslint ./src/` to check for code quality issues
-- Fix any errors or warnings before creating the release
+#### 3. Run linting and fix issues
+- Run `npx eslint ./src/` to check for code quality issues
+- **Fix all errors and warnings** before creating the release
+- Common issues to watch for:
+  - Unused imports (remove them)
+  - Floating promises (add `await` or `void` operator)
+  - Sentence case violations in UI text (use lowercase after first word)
+  - Unnecessary type assertions (remove redundant `as` casts)
+  - Misused promises in callbacks (use `void` for fire-and-forget async calls)
+  - Console statements (only `console.warn`, `console.error`, and `console.debug` are allowed)
 
 ## Security, privacy, and compliance
 
@@ -206,6 +213,7 @@ Follow Obsidian's **Developer Policies** and **Plugin Guidelines**. In particula
 - Write idempotent code paths so reload/unload doesn't leak listeners or intervals.
 - Use `this.register*` helpers for everything that needs cleanup.
 - When refreshing collapsible settings sections, preserve the expand state by setting `callbacks.setExpandState({ workflowId: workflow.id })` before calling `callbacks.refresh()` if the section is currently expanded.
+- When editing action settings that trigger a UI refresh, use `preserveActionExpandState()` callback to maintain expanded state.
 
 **Don't**
 - Introduce network calls without an obvious user-facing reason and documentation.
@@ -289,13 +297,185 @@ this.registerDomEvent(window, "resize", () => { /* ... */ });
 this.registerInterval(window.setInterval(() => { /* ... */ }, 1000));
 ```
 
+## Action-based workflow system
+
+Workflows are containers for sequential actions. Each workflow can contain multiple actions that execute in order, with later actions able to reference outputs from earlier actions using tokens.
+
+### Workflow structure
+
+```ts
+interface WorkflowConfig {
+  id: string;                    // Unique identifier
+  name: string;                  // Display name
+  actions: WorkflowAction[];     // Sequential list of actions
+  outputType: WorkflowOutputType; // 'popup' | 'new-note' | 'at-cursor'
+  outputFolder: string;          // Folder for new-note output
+}
+```
+
+### Action types
+
+There are two action types: **chat** and **transcription**.
+
+#### Chat action
+
+Sends a prompt to an AI model and receives a response.
+
+```ts
+interface ChatAction extends BaseAction {
+  type: 'chat';
+  promptText: string;           // The prompt text (when inline)
+  promptSourceType: PromptSourceType; // 'inline' | 'from-file'
+  promptFilePath: string;       // Path to prompt file (when from-file)
+  contexts?: ChatContextConfig[]; // Context sources (selection, clipboard, etc.)
+}
+```
+
+**Configuration options:**
+- **Provider**: Select which AI provider and model to use
+- **Prompt source**: Inline text or load from a file in the vault
+- **Prompt text**: The prompt template with token placeholders
+
+#### Transcription action
+
+Transcribes audio/video content using a speech-to-text model.
+
+```ts
+interface TranscriptionAction extends BaseAction {
+  type: 'transcription';
+  transcriptionContext?: {
+    mediaType: 'video' | 'audio';
+    sourceUrlToken?: string;    // Token containing the URL (e.g., 'workflow.clipboard')
+  };
+  language?: string;            // ISO language code (e.g., 'en', 'es')
+  timestampGranularity?: 'disabled' | 'segment' | 'word';
+}
+```
+
+**Configuration options:**
+- **Provider**: Select which AI provider and model to use (must support transcription)
+- **Media type**: Video or audio
+- **Source URL**: Token containing the media URL (default: `workflow.clipboard`)
+- **Language**: Optional language hint for better accuracy
+- **Timestamp granularity**: Disabled (default), segment-level, or word-level
+
+### Token system
+
+Actions can reference values from workflow context and previous action outputs using `{{tokenName}}` syntax.
+
+#### Workflow context tokens (always available)
+
+| Token | Description |
+|-------|-------------|
+| `{{workflow.selection}}` | Currently selected text in the editor |
+| `{{workflow.clipboard}}` | Contents of the system clipboard |
+| `{{workflow.file.content}}` | Full contents of the active file |
+| `{{workflow.file.path}}` | Path of the active file |
+
+#### Chat action output tokens
+
+| Token | Description |
+|-------|-------------|
+| `{{actionId.prompt}}` | The original prompt text |
+| `{{actionId.response}}` | The AI response text |
+
+#### Transcription action output tokens
+
+| Token | Description |
+|-------|-------------|
+| `{{actionId.title}}` | Video title |
+| `{{actionId.author}}` | Video uploader/author |
+| `{{actionId.sourceUrl}}` | Original video URL |
+| `{{actionId.description}}` | Video description |
+| `{{actionId.tags}}` | Video tags (comma-separated) |
+| `{{actionId.transcription}}` | Plain transcription text |
+| `{{actionId.transcriptionWithTimestamps}}` | Transcription with `[MM:SS]` prefixes (only when timestamps enabled) |
+
+### Token replacement example
+
+A workflow with two actions:
+1. **Transcription action** (id: `abc123`) - transcribes a video from clipboard URL
+2. **Chat action** - summarizes the transcription
+
+The chat action's prompt can reference the transcription output:
+
+```
+Summarize the following video transcript:
+
+Title: {{abc123.title}}
+Author: {{abc123.author}}
+
+Transcript:
+{{abc123.transcription}}
+```
+
+### Workflow execution flow
+
+```
+┌─────────────────┐     ┌─────────────────┐     ┌─────────────────┐     ┌────────────────┐
+│ Gather workflow │ ──▶ │ Execute Action 1│ ──▶ │ Execute Action 2│ ──▶ │ Output Handler │
+│ context         │     │ (store tokens)  │     │ (use prev tokens)│     │ (final result) │
+└─────────────────┘     └─────────────────┘     └─────────────────┘     └────────────────┘
+```
+
+### Settings UI expand state preservation
+
+When editing action settings that trigger a UI refresh (e.g., changing prompt source type), preserve the action's expanded state:
+
+```ts
+// In action-specific settings functions, use preserveActionExpandState callback
+.onChange(async (value) => {
+  action.someProperty = value;
+  await plugin.saveSettings();
+  preserveActionExpandState();  // Sets both workflowId and actionId
+  callbacks.refresh();
+});
+```
+
+The `ExpandOnNextRenderState` interface tracks which items should be expanded:
+
+```ts
+interface ExpandOnNextRenderState {
+  providerId?: string;
+  modelId?: string;
+  workflowId?: string;
+  actionId?: string;  // Added for action expand state
+  availableTokensExpanded?: boolean;
+}
+```
+
+### Creating new actions (avoiding shared references)
+
+When creating new actions, always create fresh instances of nested objects and arrays:
+
+```ts
+// Correct: Create new array/object instances
+const newWorkflow: WorkflowConfig = {
+  id: generateId(),
+  ...DEFAULT_WORKFLOW_CONFIG,
+  actions: []  // New array, not shared reference
+};
+
+const newChatAction = {
+  ...DEFAULT_CHAT_ACTION,
+  id: generateId(),
+  contexts: []  // New array
+};
+
+const newTranscriptionAction = {
+  ...DEFAULT_TRANSCRIPTION_ACTION,
+  id: generateId(),
+  transcriptionContext: { mediaType: 'video', sourceUrlToken: 'workflow.clipboard' }  // New object
+};
+```
+
 ## Handler architecture
 
 The plugin uses a handler-based architecture to separate concerns for workflow execution:
 
 ### Input handlers (`src/handlers/input/`)
 
-Input handlers acquire media for transcription workflows. Each handler implements the `InputHandler` interface:
+Input handlers acquire media for transcription actions. Each handler implements the `InputHandler` interface:
 
 ```ts
 interface InputHandler {
@@ -307,6 +487,7 @@ interface InputHandler {
 - `VaultFileInputHandler` - Prompts user to select an audio file from the vault
 - `ClipboardUrlInputHandler` - Extracts audio from a video URL in the clipboard (uses yt-dlp)
 - `SelectionUrlInputHandler` - Extracts audio from a video URL in the current text selection
+- `TokenUrlInputHandler` - Extracts audio from a URL resolved from a token value
 
 **InputResult structure:**
 ```ts
@@ -343,26 +524,15 @@ interface OutputHandler {
 **To add a new input handler:**
 1. Create a new file in `src/handlers/input/` (e.g., `my-custom-input-handler.ts`)
 2. Implement the `InputHandler` interface
-3. Export from `src/handlers/input/index.ts`
-4. Add the new source type to `TranscriptionSourceType` in `src/settings/types.ts`
-5. Update `createInputHandler()` in `src/processing/workflow-executor.ts`
+3. Export from `src/handlers/input/index.ts` and `src/handlers/index.ts`
 
 **To add a new output handler:**
 1. Create a new file in `src/handlers/output/` (e.g., `my-custom-output-handler.ts`)
 2. Implement the `OutputHandler` interface
-3. Export from `src/handlers/output/index.ts`
+3. Export from `src/handlers/output/index.ts` and `src/handlers/index.ts`
 4. Add the new output type to `WorkflowOutputType` in `src/settings/types.ts`
 5. Update `createOutputHandler()` in `src/processing/workflow-executor.ts`
 
-### Workflow execution flow
-
-```
-┌─────────────────┐     ┌───────────────┐     ┌────────────────┐     ┌────────────────┐
-│ User triggers   │ ──▶ │ Input Handler │ ──▶ │ AI Provider    │ ──▶ │ Output Handler │
-│ workflow        │     │ (get media)   │     │ (transcribe)   │     │ (show result)  │
-└─────────────────┘     └───────────────┘     └────────────────┘     └────────────────┘
-```
-
 ## Troubleshooting
 
 - Plugin doesn't load after build: ensure `main.js` and `manifest.json` are at the top level of the plugin folder under `<Vault>/.obsidian/plugins/<plugin-id>/`. 
diff --git a/manifest.json b/manifest.json
index de98265..c309a77 100644
--- a/manifest.json
+++ b/manifest.json
@@ -1,7 +1,7 @@
 {
 	"id": "ai-toolbox",
 	"name": "AI Toolbox",
-	"version": "1.0.3",
+	"version": "1.0.4",
 	"minAppVersion": "0.15.0",
 	"description": "Personal collection of AI tools to enhance the Obsidian.md workflow.",
 	"author": "dalinicus",
diff --git a/src/components/path-picker.ts b/src/components/path-picker.ts
index d63bc9e..b6bfef0 100644
--- a/src/components/path-picker.ts
+++ b/src/components/path-picker.ts
@@ -1,4 +1,4 @@
-import { App, Setting, AbstractInputSuggest, TFolder, TFile, TAbstractFile } from "obsidian";
+import { App, Setting, AbstractInputSuggest, TAbstractFile } from "obsidian";
 
 /**
  * Type of path selection.
diff --git a/src/handlers/index.ts b/src/handlers/index.ts
index dac0517..9a764de 100644
--- a/src/handlers/index.ts
+++ b/src/handlers/index.ts
@@ -3,7 +3,8 @@ export type { InputHandler, InputContext, InputResult } from './input';
 export {
     VaultFileInputHandler,
     ClipboardUrlInputHandler,
-    SelectionUrlInputHandler
+    SelectionUrlInputHandler,
+    TokenUrlInputHandler
 } from './input';
 
 // Output handlers
diff --git a/src/handlers/input/index.ts b/src/handlers/input/index.ts
index 9afca07..375a895 100644
--- a/src/handlers/input/index.ts
+++ b/src/handlers/input/index.ts
@@ -5,4 +5,5 @@ export type { InputHandler, InputContext, InputResult } from './types';
 export { VaultFileInputHandler } from './vault-file-input-handler';
 export { ClipboardUrlInputHandler } from './clipboard-url-input-handler';
 export { SelectionUrlInputHandler } from './selection-url-input-handler';
+export { TokenUrlInputHandler } from './token-url-input-handler';
 
diff --git a/src/handlers/input/token-url-input-handler.ts b/src/handlers/input/token-url-input-handler.ts
new file mode 100644
index 0000000..7d92a53
--- /dev/null
+++ b/src/handlers/input/token-url-input-handler.ts
@@ -0,0 +1,37 @@
+import { Notice } from 'obsidian';
+import { InputHandler, InputContext, InputResult } from './types';
+import { extractAudioFromUrl } from '../../processing';
+
+/**
+ * Input handler that extracts audio from a URL resolved from a token value.
+ * Uses yt-dlp to download and convert the video to audio.
+ */
+export class TokenUrlInputHandler implements InputHandler {
+    private url: string;
+
+    constructor(url: string) {
+        this.url = url;
+    }
+
+    async getInput(context: InputContext): Promise<InputResult | null> {
+        if (!this.url || !this.url.trim()) {
+            new Notice('No URL found in the selected token');
+            return null;
+        }
+
+        new Notice('Processing URL from token...');
+
+        const result = await extractAudioFromUrl(this.url.trim(), context.settings);
+
+        if (!result) {
+            return null;
+        }
+
+        return {
+            audioFilePath: result.audioFilePath,
+            sourceUrl: result.sourceUrl,
+            metadata: result.metadata
+        };
+    }
+}
+
diff --git a/src/logging/log-manager.ts b/src/logging/log-manager.ts
index eeb96dc..6685aa6 100644
--- a/src/logging/log-manager.ts
+++ b/src/logging/log-manager.ts
@@ -37,10 +37,8 @@ class LogManager {
                 console.warn(formattedMsg, details !== undefined ? details : '');
                 break;
             case LogLevel.DEBUG:
-                console.debug(formattedMsg, details !== undefined ? details : '');
-                break;
             default:
-                console.log(formattedMsg, details !== undefined ? details : '');
+                console.debug(formattedMsg, details !== undefined ? details : '');
         }
 
         this.notifyListeners();
@@ -98,26 +96,26 @@ class LogManager {
 export const logManager = new LogManager();
 
 // Convenience functions for logging at specific levels
-export function logDebug(category: LogCategoryType | string, message: string, details?: unknown): void {
+export function logDebug(category: LogCategoryType, message: string, details?: unknown): void {
     logManager.log(LogLevel.DEBUG, category, message, details);
 }
 
-export function logInfo(category: LogCategoryType | string, message: string, details?: unknown): void {
+export function logInfo(category: LogCategoryType, message: string, details?: unknown): void {
     logManager.log(LogLevel.INFO, category, message, details);
 }
 
-export function logWarn(category: LogCategoryType | string, message: string, details?: unknown): void {
+export function logWarn(category: LogCategoryType, message: string, details?: unknown): void {
     logManager.log(LogLevel.WARN, category, message, details);
 }
 
-export function logError(category: LogCategoryType | string, message: string, details?: unknown): void {
+export function logError(category: LogCategoryType, message: string, details?: unknown): void {
     logManager.log(LogLevel.ERROR, category, message, details);
 }
 
 /**
  * Log at INFO level and also display an Obsidian Notice to the user.
  */
-export function logNotice(category: LogCategoryType | string, message: string, details?: unknown): void {
+export function logNotice(category: LogCategoryType, message: string, details?: unknown): void {
     logManager.log(LogLevel.INFO, category, message, details);
     new Notice(message);
 }
diff --git a/src/logging/log-pane-view.ts b/src/logging/log-pane-view.ts
index ae612f0..c105809 100644
--- a/src/logging/log-pane-view.ts
+++ b/src/logging/log-pane-view.ts
@@ -25,7 +25,7 @@ export class LogPaneView extends ItemView {
     }
 
     getDisplayText(): string {
-        return 'AI Toolbox Log';
+        return 'AI toolbox log';
     }
 
     getIcon(): string {
@@ -53,7 +53,7 @@ export class LogPaneView extends ItemView {
         const levels = ['DEBUG', 'INFO', 'WARN', 'ERROR'];
         levels.forEach((level, i) => {
             const option = select.createEl('option', { value: String(i), text: level });
-            if (i === this.filterLevel) option.selected = true;
+            if (i === (this.filterLevel as number)) option.selected = true;
         });
         select.addEventListener('change', () => {
             this.filterLevel = parseInt(select.value);
@@ -77,7 +77,7 @@ export class LogPaneView extends ItemView {
         clearBtn.addEventListener('click', () => logManager.clear());
 
         // Copy button
-        const copyBtn = controls.createEl('button', { text: 'Copy All' });
+        const copyBtn = controls.createEl('button', { text: 'Copy all' });
         copyBtn.addEventListener('click', () => this.copyLogs());
     }
 
@@ -105,7 +105,7 @@ export class LogPaneView extends ItemView {
         const text = logManager.getLogs()
             .map(e => `[${e.timestamp.toISOString()}] [${LogLevel[e.level]}] [${e.category}] ${e.message}`)
             .join('\n');
-        navigator.clipboard.writeText(text);
+        void navigator.clipboard.writeText(text);
     }
 
     async onClose(): Promise<void> {
diff --git a/src/main.ts b/src/main.ts
index b0b7883..f1d6132 100644
--- a/src/main.ts
+++ b/src/main.ts
@@ -14,7 +14,7 @@ export default class AIToolboxPlugin extends Plugin {
 		this.registerView(VIEW_TYPE_LOG, (leaf) => new LogPaneView(leaf, this));
 
 		// Add ribbon icon for log view
-		this.addRibbonIcon('scroll-text', 'Show AI Toolbox Log', () => {
+		this.addRibbonIcon('scroll-text', 'Show AI toolbox log', () => {
 			void this.activateLogView();
 		});
 
@@ -41,8 +41,7 @@ export default class AIToolboxPlugin extends Plugin {
 	}
 
 	onunload() {
-		// Clean up log view leaves
-		this.app.workspace.detachLeavesOfType(VIEW_TYPE_LOG);
+		// Log views are cleaned up automatically by Obsidian when the plugin unloads
 	}
 
 	/**
@@ -61,7 +60,7 @@ export default class AIToolboxPlugin extends Plugin {
 		}
 
 		if (leaf) {
-			workspace.revealLeaf(leaf);
+			await workspace.revealLeaf(leaf);
 		}
 	}
 
diff --git a/src/processing/action-executor.ts b/src/processing/action-executor.ts
index b212751..f442a59 100644
--- a/src/processing/action-executor.ts
+++ b/src/processing/action-executor.ts
@@ -1,20 +1,16 @@
 import { App, TFile } from 'obsidian';
-import { AIToolboxSettings, ChatAction, TranscriptionAction, WorkflowAction, TranscriptionSourceType } from '../settings';
+import { AIToolboxSettings, ChatAction, TranscriptionAction, WorkflowAction } from '../settings';
 import { createActionProvider, ChatMessage, TranscriptionOptions } from '../providers';
 import {
-    InputHandler,
     InputContext,
     InputResult,
-    VaultFileInputHandler,
-    ClipboardUrlInputHandler,
-    SelectionUrlInputHandler
+    TokenUrlInputHandler
 } from '../handlers';
 import {
     createChatWorkflowTokens,
     createTranscriptionWorkflowTokens,
-    gatherContextValues,
-    replaceContextTokens,
-    hasContextTokens
+    ContextTokenValues,
+    replaceWorkflowContextTokens
 } from './workflow-chaining';
 import { logInfo, logDebug, logNotice, LogCategory } from '../logging';
 
@@ -56,6 +52,8 @@ export interface ActionExecutionContext {
     dependencyResults: Map<string, ActionExecutionResult>;
     /** The workflow name (for logging) */
     workflowName: string;
+    /** Workflow-level context values (clipboard, selection, etc.) gathered at workflow start */
+    workflowContext: ContextTokenValues;
 }
 
 /**
@@ -84,19 +82,43 @@ export function replaceActionTokens(
 }
 
 /**
- * Create an input handler based on the transcription source type.
+ * Resolve a token value from the execution context.
+ * Tokens can reference workflow context (e.g., 'workflow.clipboard') or
+ * previous action results (e.g., 'chat1.response').
  */
-function createInputHandler(sourceType: TranscriptionSourceType): InputHandler {
-    switch (sourceType) {
-        case 'select-file-from-vault':
-            return new VaultFileInputHandler();
-        case 'url-from-clipboard':
-            return new ClipboardUrlInputHandler();
-        case 'url-from-selection':
-            return new SelectionUrlInputHandler();
-        default:
-            return new VaultFileInputHandler();
+function resolveTokenValue(
+    tokenName: string,
+    context: ActionExecutionContext
+): string | undefined {
+    // Check if it's a workflow context token
+    if (tokenName.startsWith('workflow.')) {
+        const contextKey = tokenName.substring('workflow.'.length);
+        switch (contextKey) {
+            case 'selection':
+                return context.workflowContext.selection;
+            case 'clipboard':
+                return context.workflowContext.clipboard;
+            case 'file.content':
+                return context.workflowContext.fileContent;
+            case 'file.path':
+                return context.workflowContext.filePath;
+            default:
+                return undefined;
+        }
+    }
+
+    // Check if it's a previous action token (e.g., 'chat1.response')
+    const dotIndex = tokenName.indexOf('.');
+    if (dotIndex > 0) {
+        const actionId = tokenName.substring(0, dotIndex);
+        const tokenKey = tokenName.substring(dotIndex + 1);
+        const actionResult = context.previousResults.get(actionId);
+        if (actionResult) {
+            return actionResult.tokens[tokenKey];
+        }
     }
+
+    return undefined;
 }
 
 /**
@@ -171,11 +193,8 @@ export async function executeChatAction(
         promptText = replaceActionTokens(promptText, context.dependencyResults);
     }
 
-    // Replace context tokens ({{selection}}, {{clipboard}}, etc.)
-    if (hasContextTokens(promptText)) {
-        const contextValues = await gatherContextValues(context.app);
-        promptText = replaceContextTokens(promptText, contextValues);
-    }
+    // Replace workflow context tokens ({{workflow.selection}}, {{workflow.clipboard}}, etc.)
+    promptText = replaceWorkflowContextTokens(promptText, context.workflowContext);
 
     if (!promptText.trim()) {
         return { ...baseResult, error: 'Empty prompt text' };
@@ -229,8 +248,15 @@ export async function executeTranscriptionAction(
         return { ...baseResult, error: 'Provider does not support transcription' };
     }
 
-    const sourceType: TranscriptionSourceType = action.transcriptionContext?.sourceType ?? 'url-from-clipboard';
-    const inputHandler = createInputHandler(sourceType);
+    // Resolve the source URL from the configured token
+    const sourceUrlToken = action.transcriptionContext?.sourceUrlToken ?? 'workflow.clipboard';
+    const sourceUrl = resolveTokenValue(sourceUrlToken, context);
+
+    if (!sourceUrl || !sourceUrl.trim()) {
+        return { ...baseResult, error: `No URL found in token {{${sourceUrlToken}}}` };
+    }
+
+    const inputHandler = new TokenUrlInputHandler(sourceUrl);
 
     // Create a minimal workflow-like object for InputContext compatibility
     const inputContext: InputContext = {
@@ -250,7 +276,7 @@ export async function executeTranscriptionAction(
 
         const inputResult = await inputHandler.getInput(inputContext);
         if (!inputResult) {
-            return { ...baseResult, error: 'No input provided or cancelled' };
+            return { ...baseResult, error: 'Failed to extract audio from URL' };
         }
 
         logNotice(LogCategory.TRANSCRIPTION, `Transcribing audio...`);
diff --git a/src/processing/video-processor.ts b/src/processing/video-processor.ts
index cd9ed24..decbe80 100644
--- a/src/processing/video-processor.ts
+++ b/src/processing/video-processor.ts
@@ -14,6 +14,7 @@ export interface VideoProcessorConfig {
     ytdlpLocation?: string;
     ffmpegLocation?: string;
     impersonateBrowser: string;
+    cookiesFromBrowser?: string;
     keepVideo: boolean;
     outputDirectory?: string;
 }
@@ -108,6 +109,10 @@ function spawnYtDlp(
             args.push('--ffmpeg-location', config.ffmpegLocation);
         }
 
+        if (config.cookiesFromBrowser) {
+            args.push('--cookies-from-browser', config.cookiesFromBrowser);
+        }
+
         args.push(
             '--impersonate', config.impersonateBrowser,
             '-o', outputTemplate,
@@ -205,6 +210,7 @@ function settingsToProcessorConfig(settings: AIToolboxSettings): VideoProcessorC
         ytdlpLocation: settings.ytdlpLocation,
         ffmpegLocation: settings.ffmpegLocation,
         impersonateBrowser: settings.impersonateBrowser,
+        cookiesFromBrowser: settings.cookiesFromBrowser,
         keepVideo: settings.keepVideo,
         outputDirectory: settings.outputDirectory,
     };
diff --git a/src/processing/workflow-chaining.ts b/src/processing/workflow-chaining.ts
index a3de469..5b28d93 100644
--- a/src/processing/workflow-chaining.ts
+++ b/src/processing/workflow-chaining.ts
@@ -165,16 +165,16 @@ export interface ContextTokenValues {
     /** The currently selected text in the editor */
     selection?: string;
     /** The full contents of the active file */
-    activeTabContent?: string;
-    /** The filename of the active file */
-    activeTabFilename?: string;
+    fileContent?: string;
+    /** The path of the active file */
+    filePath?: string;
     /** The clipboard contents */
     clipboard?: string;
 }
 
 /**
  * Gather context values from the current editor/clipboard state.
- * This retrieves selection, active tab content, filename, and clipboard for token replacement.
+ * This retrieves selection, file content, file path, and clipboard for token replacement.
  */
 export async function gatherContextValues(app: App): Promise<ContextTokenValues> {
     const values: ContextTokenValues = {};
@@ -190,9 +190,9 @@ export async function gatherContextValues(app: App): Promise<ContextTokenValues>
 
         const file = activeView.file;
         if (file) {
-            values.activeTabFilename = file.name;
+            values.filePath = file.path;
             try {
-                values.activeTabContent = await app.vault.read(file);
+                values.fileContent = await app.vault.read(file);
             } catch {
                 // Ignore read errors
             }
@@ -212,24 +212,27 @@ export async function gatherContextValues(app: App): Promise<ContextTokenValues>
 }
 
 /**
- * Replace context tokens in a prompt with actual values.
- * Handles simple tokens like {{selection}}, {{clipboard}}, {{activeTabContent}}, {{activeTabFilename}}.
+ * Replace workflow context tokens in a prompt with actual values.
+ * Handles tokens like {{workflow.selection}}, {{workflow.clipboard}},
+ * {{workflow.file.content}}, {{workflow.file.path}}.
+ *
+ * These tokens are gathered once at workflow start and shared across all actions.
  */
-export function replaceContextTokens(
+export function replaceWorkflowContextTokens(
     promptText: string,
     values: ContextTokenValues
 ): string {
-    // Match simple tokens like {{selection}} (no dot, simple alphanumeric name)
-    const tokenPattern = /\{\{([a-zA-Z]+)\}\}/g;
+    // Match workflow context tokens like {{workflow.selection}} or {{workflow.file.content}}
+    const tokenPattern = /\{\{workflow\.([a-zA-Z.]+)\}\}/g;
 
     return promptText.replace(tokenPattern, (match, tokenName: string) => {
         switch (tokenName) {
             case 'selection':
                 return values.selection ?? match;
-            case 'activeTabContent':
-                return values.activeTabContent ?? match;
-            case 'activeTabFilename':
-                return values.activeTabFilename ?? match;
+            case 'file.content':
+                return values.fileContent ?? match;
+            case 'file.path':
+                return values.filePath ?? match;
             case 'clipboard':
                 return values.clipboard ?? match;
             default:
@@ -239,11 +242,3 @@ export function replaceContextTokens(
     });
 }
 
-/**
- * Check if a prompt contains any context tokens that need replacement.
- */
-export function hasContextTokens(promptText: string): boolean {
-    const contextTokens = ['selection', 'activeTabContent', 'activeTabFilename', 'clipboard'];
-    return contextTokens.some(token => promptText.includes(`{{${token}}}`));
-}
-
diff --git a/src/processing/workflow-executor.ts b/src/processing/workflow-executor.ts
index a5878dd..40481ab 100644
--- a/src/processing/workflow-executor.ts
+++ b/src/processing/workflow-executor.ts
@@ -7,15 +7,16 @@ import {
     OutputContext,
     NewNoteOutputHandler,
     AtCursorOutputHandler,
-    PopupOutputHandler,
-    InputResult
+    PopupOutputHandler
 } from '../handlers';
 import {
     WorkflowExecutionResult,
     WorkflowResultsMap,
     detectCircularDependency,
     hasWorkflowDependencies,
-    getDependencyWorkflowIds
+    getDependencyWorkflowIds,
+    gatherContextValues,
+    ContextTokenValues
 } from './workflow-chaining';
 import {
     ActionExecutionResult,
@@ -195,6 +196,9 @@ export async function executeWorkflow(
         });
     }
 
+    // Gather workflow context values once at the start (clipboard, selection, etc.)
+    const workflowContext: ContextTokenValues = await gatherContextValues(app);
+
     // Execute actions sequentially
     const actionResults: ActionResultsMap = new Map();
     let lastResult: ActionExecutionResult | null = null;
@@ -207,7 +211,8 @@ export async function executeWorkflow(
             settings,
             previousResults: actionResults,
             dependencyResults: dependencyActionResults,
-            workflowName: workflow.name
+            workflowName: workflow.name,
+            workflowContext
         };
 
         const result = await executeAction(action, context);
@@ -263,6 +268,9 @@ async function executeWorkflowInternal(
         return { ...baseResult, error: 'No actions configured' };
     }
 
+    // Gather workflow context values once at the start
+    const workflowContext: ContextTokenValues = await gatherContextValues(app);
+
     // Execute actions sequentially
     const actionResults: ActionResultsMap = new Map();
 
@@ -272,7 +280,8 @@ async function executeWorkflowInternal(
             settings,
             previousResults: actionResults,
             dependencyResults: new Map(),
-            workflowName: workflow.name
+            workflowName: workflow.name,
+            workflowContext
         };
 
         const result = await executeAction(action, context);
diff --git a/src/settings/additional-settings.ts b/src/settings/additional-settings.ts
index 9e413f8..18ead84 100644
--- a/src/settings/additional-settings.ts
+++ b/src/settings/additional-settings.ts
@@ -41,6 +41,25 @@ export function displayAdditionalSettings(
 				await plugin.saveSettings();
 			}));
 
+	new Setting(containerEl)
+		.setName('Use cookies from browser')
+		.setDesc('Extract cookies from a browser for authentication (required for some age-restricted or private content)')
+		.addDropdown(dropdown => dropdown
+			.addOption('', 'None')
+			.addOption('chrome', 'Chrome')
+			.addOption('edge', 'Edge')
+			.addOption('safari', 'Safari')
+			.addOption('firefox', 'Firefox')
+			.addOption('brave', 'Brave')
+			.addOption('chromium', 'Chromium')
+			.addOption('opera', 'Opera')
+			.addOption('vivaldi', 'Vivaldi')
+			.setValue(plugin.settings.cookiesFromBrowser)
+			.onChange(async (value) => {
+				plugin.settings.cookiesFromBrowser = value;
+				await plugin.saveSettings();
+			}));
+
 	const ytdlpPathSetting = new Setting(containerEl)
 		.setName('yt-dlp path') // eslint-disable-line obsidianmd/ui/sentence-case -- proper noun
 		.setDesc('Path to yt-dlp binary directory (leave empty to use system PATH)') // eslint-disable-line obsidianmd/ui/sentence-case -- proper noun
diff --git a/src/settings/index.ts b/src/settings/index.ts
index d388a89..290131a 100644
--- a/src/settings/index.ts
+++ b/src/settings/index.ts
@@ -24,7 +24,6 @@ export type {
 	WorkflowType,
 	PromptSourceType,
 	TranscriptionMediaType,
-	TranscriptionSourceType,
 	TranscriptionContextConfig,
 	TimestampGranularity,
 	ChatContextType,
diff --git a/src/settings/types.ts b/src/settings/types.ts
index f5807c9..ecd9cd2 100644
--- a/src/settings/types.ts
+++ b/src/settings/types.ts
@@ -60,11 +60,6 @@ export type WorkflowType = 'chat' | 'transcription';
  */
 export type TranscriptionMediaType = 'video' | 'audio';
 
-/**
- * Source type for transcription input
- */
-export type TranscriptionSourceType = 'select-file-from-vault' | 'url-from-clipboard' | 'url-from-selection';
-
 /**
  * Timestamp granularity for transcription output.
  * - 'disabled': No timestamps (uses simpler JSON format, reduces token usage)
@@ -78,7 +73,8 @@ export type TimestampGranularity = 'disabled' | 'segment' | 'word';
  */
 export interface TranscriptionContextConfig {
 	mediaType: TranscriptionMediaType;
-	sourceType: TranscriptionSourceType;
+	/** Token name to resolve for the source URL (e.g., 'workflow.clipboard', 'chat1.response') */
+	sourceUrlToken?: string;
 }
 
 /**
@@ -170,7 +166,7 @@ export const DEFAULT_TRANSCRIPTION_ACTION: Omit<TranscriptionAction, 'id'> = {
 	type: 'transcription',
 	name: 'Transcription',
 	provider: null,
-	transcriptionContext: { mediaType: 'video', sourceType: 'url-from-clipboard' },
+	transcriptionContext: { mediaType: 'video', sourceUrlToken: 'workflow.clipboard' },
 	language: '',
 	timestampGranularity: 'disabled'
 };
@@ -203,6 +199,7 @@ export const DEFAULT_WORKFLOW_CONFIG: Omit<WorkflowConfig, 'id'> = {
 
 export interface AIToolboxSettings {
 	impersonateBrowser: string;
+	cookiesFromBrowser: string;
 	ytdlpLocation: string;
 	ffmpegLocation: string;
 	outputDirectory: string;
@@ -222,6 +219,7 @@ export function generateId(): string {
 
 export const DEFAULT_SETTINGS: AIToolboxSettings = {
 	impersonateBrowser: 'chrome',
+	cookiesFromBrowser: '',
 	ytdlpLocation: '',
 	ffmpegLocation: '',
 	outputDirectory: '',
@@ -242,6 +240,7 @@ export interface ExpandOnNextRenderState {
 	providerId?: string;
 	modelId?: string;
 	workflowId?: string;
+	actionId?: string;
 	availableTokensExpanded?: boolean;
 }
 
diff --git a/src/settings/workflows.ts b/src/settings/workflows.ts
index ce49730..ae16192 100644
--- a/src/settings/workflows.ts
+++ b/src/settings/workflows.ts
@@ -9,9 +9,7 @@ import {
 	ActionType,
 	PromptSourceType,
 	TranscriptionMediaType,
-	TranscriptionSourceType,
 	TimestampGranularity,
-	ChatContextType,
 	ExpandOnNextRenderState,
 	generateId,
 	DEFAULT_WORKFLOW_CONFIG,
@@ -21,10 +19,12 @@ import {
 import { createCollapsibleSection } from "../components/collapsible-section";
 import { createPathPicker } from "../components/path-picker";
 import {
-	CHAT_CONTEXT_TYPE_LABELS,
-	CHAT_CONTEXT_TYPE_DESCRIPTIONS,
-	getAvailableContextTypes
-} from "../handlers";
+	getAvailableTokensForAction,
+	TokenGroup,
+	TokenDefinition,
+	generateGroupTemplate
+} from "../tokens";
+import { setIcon, Notice } from "obsidian";
 
 /**
  * Callbacks for the workflows settings tab to communicate with the main settings tab
@@ -71,7 +71,8 @@ export function displayWorkflowsSettings(
 			.onClick(async () => {
 				const newWorkflow: WorkflowConfig = {
 					id: generateId(),
-					...DEFAULT_WORKFLOW_CONFIG
+					...DEFAULT_WORKFLOW_CONFIG,
+					actions: []  // Create new array to avoid shared reference
 				};
 				plugin.settings.workflows.push(newWorkflow);
 				callbacks.setExpandState({ workflowId: newWorkflow.id });
@@ -165,9 +166,9 @@ function displayWorkflowSettings(
 			description: 'Folder where new notes will be created',
 			initialPath: workflow.outputFolder ?? '',
 			allowFiles: false,
-			onChange: async (path: string) => {
+			onChange: (path: string) => {
 				workflow.outputFolder = path;
-				await plugin.saveSettings();
+				void plugin.saveSettings();
 			}
 		});
 	}
@@ -214,13 +215,16 @@ function displayActionsSection(
 					newAction = {
 						...DEFAULT_CHAT_ACTION,
 						id: generateId(),
-						name: `Chat ${workflow.actions.length + 1}`
+						name: `Chat ${workflow.actions.length + 1}`,
+						contexts: []  // Create new array to avoid shared reference
 					};
 				} else {
 					newAction = {
 						...DEFAULT_TRANSCRIPTION_ACTION,
 						id: generateId(),
-						name: `Transcription ${workflow.actions.length + 1}`
+						name: `Transcription ${workflow.actions.length + 1}`,
+						// Create new object to avoid shared reference
+						transcriptionContext: { mediaType: 'video', sourceUrlToken: 'workflow.clipboard' }
 					};
 				}
 
@@ -254,14 +258,16 @@ function displayActionSettings(
 	isExpanded: () => boolean
 ): void {
 	const actionIcon = action.type === 'transcription' ? 'audio-lines' : 'message-circle';
+	const expandState = callbacks.getExpandState();
+	const shouldExpandAction = expandState.workflowId === workflow.id && expandState.actionId === action.id;
 
-	const { contentContainer, updateTitle } = createCollapsibleSection({
+	const { contentContainer, updateTitle, isExpanded: isActionExpanded } = createCollapsibleSection({
 		containerEl,
 		title: action.name || `Action ${index + 1}`,
 		containerClass: 'action-container',
 		contentClass: 'action-content',
 		headerClass: 'action-header',
-		startExpanded: false,
+		startExpanded: shouldExpandAction,
 		isHeading: false,
 		icon: actionIcon,
 		onDelete: async () => {
@@ -274,6 +280,20 @@ function displayActionSettings(
 		},
 	});
 
+	// Helper to preserve action expand state on refresh
+	const preserveActionExpandState = () => {
+		if (isExpanded() && isActionExpanded()) {
+			callbacks.setExpandState({ workflowId: workflow.id, actionId: action.id });
+		} else if (isExpanded()) {
+			callbacks.setExpandState({ workflowId: workflow.id });
+		}
+	};
+
+	// Clear the action expand state after applying it
+	if (shouldExpandAction) {
+		callbacks.setExpandState({ workflowId: workflow.id });
+	}
+
 	// Action name
 	new Setting(contentContainer)
 		.setName('Name')
@@ -286,11 +306,102 @@ function displayActionSettings(
 				await plugin.saveSettings();
 			}));
 
+	// Available tokens section (for chat actions only, since they use prompts)
+	if (action.type === 'chat') {
+		displayAvailableTokensSection(contentContainer, workflow, index, plugin);
+	}
+
 	// Route to type-specific settings
 	if (action.type === 'transcription') {
-		displayTranscriptionActionSettings(contentContainer, plugin, action as TranscriptionAction, callbacks, isExpanded, workflow);
+		displayTranscriptionActionSettings(contentContainer, plugin, action, callbacks, workflow, preserveActionExpandState);
 	} else {
-		displayChatActionSettings(contentContainer, plugin, action as ChatAction, callbacks, isExpanded, workflow);
+		displayChatActionSettings(contentContainer, plugin, action, callbacks, preserveActionExpandState);
+	}
+}
+
+/**
+ * Display the available tokens section for an action
+ */
+function displayAvailableTokensSection(
+	containerEl: HTMLElement,
+	workflow: WorkflowConfig,
+	actionIndex: number,
+	_plugin: AIToolboxPlugin
+): void {
+	const tokenGroups = getAvailableTokensForAction(workflow, actionIndex);
+
+	// Create collapsible section
+	const { contentContainer } = createCollapsibleSection({
+		containerEl,
+		title: 'Available Tokens',
+		containerClass: 'available-tokens-section',
+		contentClass: 'available-tokens-content',
+		headerClass: 'available-tokens-header',
+		startExpanded: false,
+		isHeading: false
+	});
+
+	// Display each token group
+	for (const group of tokenGroups) {
+		displayTokenGroup(contentContainer, group);
+	}
+}
+
+/**
+ * Copy text to clipboard and show a notification
+ */
+async function copyToClipboard(text: string, description: string): Promise<void> {
+	try {
+		await navigator.clipboard.writeText(text);
+		new Notice(`Copied ${description} to clipboard`);
+	} catch {
+		new Notice('Failed to copy to clipboard');
+	}
+}
+
+/**
+ * Display a group of tokens
+ */
+function displayTokenGroup(containerEl: HTMLElement, group: TokenGroup): void {
+	const groupEl = containerEl.createDiv('available-tokens-group');
+
+	// Group header with copy functionality
+	const headerEl = groupEl.createDiv({ cls: 'available-tokens-group-header is-clickable' });
+	headerEl.createSpan({ text: group.name });
+
+	const copyIconEl = headerEl.createSpan({ cls: 'available-tokens-copy-icon' });
+	setIcon(copyIconEl, 'copy');
+
+	headerEl.addEventListener('click', () => {
+		const template = generateGroupTemplate(group);
+		void copyToClipboard(template, group.name);
+	});
+	headerEl.setAttribute('aria-label', `Copy all ${group.name} tokens`);
+
+	displayTokenList(groupEl, group.tokens);
+}
+
+/**
+ * Display a list of tokens with click-to-copy functionality
+ */
+function displayTokenList(containerEl: HTMLElement, tokens: TokenDefinition[]): void {
+	const tokenListEl = containerEl.createEl('ul', { cls: 'available-tokens-list' });
+
+	for (const token of tokens) {
+		const tokenEl = tokenListEl.createEl('li', { cls: 'available-tokens-item' });
+
+		const tokenNameEl = tokenEl.createSpan({ cls: 'available-tokens-reference' });
+		tokenNameEl.textContent = `{{${token.name}}}`;
+		tokenNameEl.setAttribute('aria-label', 'Click to copy');
+		tokenNameEl.addEventListener('click', (e) => {
+			e.stopPropagation();
+			void copyToClipboard(`{{${token.name}}}`, 'token');
+		});
+
+		tokenEl.appendText(' ');
+
+		const tokenDescEl = tokenEl.createSpan({ cls: 'available-tokens-description' });
+		tokenDescEl.textContent = `— ${token.description}`;
 	}
 }
 
@@ -302,15 +413,11 @@ function displayChatActionSettings(
 	plugin: AIToolboxPlugin,
 	action: ChatAction,
 	callbacks: WorkflowSettingsCallbacks,
-	isExpanded: () => boolean,
-	workflow: WorkflowConfig
+	preserveActionExpandState: () => void
 ): void {
 	// Provider selection
 	displayActionProviderSelection(containerEl, plugin, action, 'chat');
 
-	// Context section
-	displayActionContextSection(containerEl, plugin, action, callbacks, isExpanded, workflow);
-
 	// Prompt source type dropdown
 	const promptSourceType = action.promptSourceType ?? 'inline';
 	new Setting(containerEl)
@@ -322,9 +429,7 @@ function displayChatActionSettings(
 			.onChange(async (value) => {
 				action.promptSourceType = value as PromptSourceType;
 				await plugin.saveSettings();
-				if (isExpanded()) {
-					callbacks.setExpandState({ workflowId: workflow.id });
-				}
+				preserveActionExpandState();
 				callbacks.refresh();
 			}));
 
@@ -356,9 +461,9 @@ function displayChatActionSettings(
 			placeholder: 'Search for file...',
 			initialPath: action.promptFilePath ?? '',
 			allowFiles: true,
-			onChange: async (path: string) => {
+			onChange: (path: string) => {
 				action.promptFilePath = path;
-				await plugin.saveSettings();
+				void plugin.saveSettings();
 			}
 		});
 	}
@@ -372,15 +477,15 @@ function displayTranscriptionActionSettings(
 	plugin: AIToolboxPlugin,
 	action: TranscriptionAction,
 	callbacks: WorkflowSettingsCallbacks,
-	isExpanded: () => boolean,
-	workflow: WorkflowConfig
+	workflow: WorkflowConfig,
+	preserveActionExpandState: () => void
 ): void {
 	// Provider selection
 	displayActionProviderSelection(containerEl, plugin, action, 'transcription');
 
 	// Ensure transcriptionContext exists
 	if (!action.transcriptionContext) {
-		action.transcriptionContext = { mediaType: 'video', sourceType: 'url-from-clipboard' };
+		action.transcriptionContext = { mediaType: 'video', sourceUrlToken: 'workflow.clipboard' };
 	}
 
 	// Media type dropdown
@@ -396,29 +501,35 @@ function displayTranscriptionActionSettings(
 			.setValue(action.transcriptionContext?.mediaType ?? 'video')
 			.onChange(async (value) => {
 				if (!action.transcriptionContext) {
-					action.transcriptionContext = { mediaType: 'video', sourceType: 'url-from-clipboard' };
+					action.transcriptionContext = { mediaType: 'video', sourceUrlToken: 'workflow.clipboard' };
 				}
 				action.transcriptionContext.mediaType = value as TranscriptionMediaType;
 				await plugin.saveSettings();
 			}));
 
-	// Source type dropdown
-	const sourceTypeOptions: Record<TranscriptionSourceType, string> = {
-		'select-file-from-vault': 'Select file from vault',
-		'url-from-clipboard': 'URL from clipboard',
-		'url-from-selection': 'URL from selection'
-	};
+	// Source URL token picker
+	const actionIndex = workflow.actions.findIndex(a => a.id === action.id);
+	const tokenGroups = getAvailableTokensForAction(workflow, actionIndex);
+
+	// Build dropdown options from available tokens
+	const tokenOptions: Record<string, string> = {};
+	for (const group of tokenGroups) {
+		for (const token of group.tokens) {
+			tokenOptions[token.name] = token.name;
+		}
+	}
+
 	new Setting(containerEl)
-		.setName('Source')
-		.setDesc('Where to get the media file from')
+		.setName('Source URL')
+		.setDesc('Select a token containing the video/audio URL')
 		.addDropdown(dropdown => dropdown
-			.addOptions(sourceTypeOptions)
-			.setValue(action.transcriptionContext?.sourceType ?? 'url-from-clipboard')
+			.addOptions(tokenOptions)
+			.setValue(action.transcriptionContext?.sourceUrlToken ?? 'workflow.clipboard')
 			.onChange(async (value) => {
 				if (!action.transcriptionContext) {
-					action.transcriptionContext = { mediaType: 'video', sourceType: 'url-from-clipboard' };
+					action.transcriptionContext = { mediaType: 'video', sourceUrlToken: 'workflow.clipboard' };
 				}
-				action.transcriptionContext.sourceType = value as TranscriptionSourceType;
+				action.transcriptionContext.sourceUrlToken = value;
 				await plugin.saveSettings();
 			}));
 
@@ -454,9 +565,7 @@ function displayTranscriptionActionSettings(
 			.onChange(async (value) => {
 				action.timestampGranularity = value as TimestampGranularity;
 				await plugin.saveSettings();
-				if (isExpanded()) {
-					callbacks.setExpandState({ workflowId: workflow.id });
-				}
+				preserveActionExpandState();
 				callbacks.refresh();
 			}));
 	granularitySetting.settingEl.toggleClass('settings-advanced-hidden', !showAdvanced);
@@ -514,75 +623,3 @@ function displayActionProviderSelection(
 				await plugin.saveSettings();
 			}));
 }
-
-/**
- * Display context section for chat actions
- */
-function displayActionContextSection(
-	containerEl: HTMLElement,
-	plugin: AIToolboxPlugin,
-	action: ChatAction,
-	callbacks: WorkflowSettingsCallbacks,
-	isExpanded: () => boolean,
-	workflow: WorkflowConfig
-): void {
-	// Ensure contexts array exists
-	if (!action.contexts) {
-		action.contexts = [];
-	}
-
-	// Helper to check if a context type is enabled
-	const isContextEnabled = (type: ChatContextType): boolean => {
-		return action.contexts?.some(c => c.type === type) ?? false;
-	};
-
-	// Helper to toggle a context type
-	const toggleContext = async (type: ChatContextType, enabled: boolean): Promise<void> => {
-		if (!action.contexts) {
-			action.contexts = [];
-		}
-
-		if (enabled) {
-			if (!action.contexts.some(c => c.type === type)) {
-				action.contexts.push({
-					id: generateId(),
-					type
-				});
-			}
-		} else {
-			action.contexts = action.contexts.filter(c => c.type !== type);
-		}
-
-		await plugin.saveSettings();
-
-		if (isExpanded()) {
-			callbacks.setExpandState({ workflowId: workflow.id });
-		}
-		callbacks.refresh();
-	};
-
-	// Context section container
-	const contextSection = containerEl.createDiv('context-section');
-	contextSection.createDiv({ text: 'Context', cls: 'context-section-title' });
-
-	// Toggles row
-	const togglesRow = contextSection.createDiv('context-toggles-row');
-
-	// Add labeled toggle for each context type
-	const availableTypes = getAvailableContextTypes();
-	for (const contextType of availableTypes) {
-		const toggleContainer = togglesRow.createDiv('context-toggle-container');
-		toggleContainer.setAttribute('aria-label', CHAT_CONTEXT_TYPE_DESCRIPTIONS[contextType]);
-
-		const label = toggleContainer.createSpan('context-toggle-label');
-		label.textContent = CHAT_CONTEXT_TYPE_LABELS[contextType];
-
-		const toggleWrapper = toggleContainer.createDiv('context-toggle-wrapper');
-		new Setting(toggleWrapper)
-			.addToggle(toggle => toggle
-				.setValue(isContextEnabled(contextType))
-				.onChange(async (value) => {
-					await toggleContext(contextType, value);
-				}));
-	}
-}
diff --git a/src/tokens/types.ts b/src/tokens/types.ts
index ffdb010..962cd64 100644
--- a/src/tokens/types.ts
+++ b/src/tokens/types.ts
@@ -1,4 +1,4 @@
-import { ActionType, TimestampGranularity } from "../settings/types";
+import { ActionType, TimestampGranularity, WorkflowConfig } from "../settings/types";
 
 /**
  * Token definition for display in the settings UI
@@ -48,6 +48,16 @@ export interface TranscriptionWorkflowTokens {
 	tags: string;
 }
 
+/**
+ * Token definitions for workflow context (available to all actions)
+ */
+export const WORKFLOW_CONTEXT_TOKENS: TokenDefinition[] = [
+	{ name: 'workflow.selection', description: 'The currently selected text in the editor' },
+	{ name: 'workflow.clipboard', description: 'The current contents of the system clipboard' },
+	{ name: 'workflow.file.content', description: 'The full contents of the active file' },
+	{ name: 'workflow.file.path', description: 'The path of the active file' }
+];
+
 /**
  * Token definitions for chat workflows (ordered for template generation)
  */
@@ -69,21 +79,6 @@ export const TRANSCRIPTION_WORKFLOW_TOKENS: TokenDefinition[] = [
 	{ name: 'transcriptionWithTimestamps', description: 'The transcription with [MM:SS] timestamps' }
 ];
 
-/**
- * Human-readable labels for token names (used in template generation)
- */
-const TOKEN_LABELS: Record<string, string> = {
-	prompt: 'Prompt',
-	response: 'Response',
-	title: 'Title',
-	author: 'Author',
-	sourceUrl: 'Source URL',
-	description: 'Description',
-	tags: 'Tags',
-	transcription: 'Transcription',
-	transcriptionWithTimestamps: 'Transcription (with timestamps)'
-};
-
 /**
  * Options for getting token definitions
  */
@@ -135,46 +130,62 @@ export function getActionTokens(
 }
 
 /**
- * Get token definitions for a workflow used as a context source.
- * Returns tokens from the last action, prefixed with the workflow ID.
- *
- * @param workflowId - The ID of the workflow
- * @param lastActionType - The type of the last action in the workflow
- * @param options - Optional settings like timestampGranularity
+ * Token group for organizing available tokens in the UI
  */
-export function getWorkflowContextTokens(
-	workflowId: string,
-	lastActionType: ActionType,
-	options?: TokenDefinitionOptions
-): TokenDefinition[] {
-	const baseTokens = getTokenDefinitionsForActionType(lastActionType, options);
+export interface TokenGroup {
+	/** Display name for this group */
+	name: string;
+	/** Tokens in this group */
+	tokens: TokenDefinition[];
+}
 
-	return baseTokens.map(token => ({
-		name: `${workflowId}.${token.name}`,
-		description: token.description
-	}));
+/**
+ * Generate a formatted template string for a token group.
+ * Format:
+ * Group Name
+ * - label: {{tokenName}}
+ */
+export function generateGroupTemplate(group: TokenGroup): string {
+	const lines = [group.name];
+	for (const token of group.tokens) {
+		const tokenName = token.name.split('.').pop() || token.name;
+		lines.push(`- ${tokenName}: {{${token.name}}}`);
+	}
+	return lines.join('\n');
 }
 
 /**
- * Generate a formatted template string containing all tokens for a workflow.
- * Used for copying a complete set of token references to the clipboard.
+ * Get all available tokens for a specific action in a workflow.
+ * Returns tokens grouped by source: workflow context and each previous action as its own group.
  *
- * @param workflowId - The ID of the workflow
- * @param lastActionType - The type of the last action in the workflow
- * @param options - Optional settings like timestampGranularity
+ * @param workflow - The workflow containing the action
+ * @param actionIndex - The index of the action in the workflow
  */
-export function generateWorkflowTokenTemplate(
-	workflowId: string,
-	lastActionType: ActionType,
-	options?: TokenDefinitionOptions
-): string {
-	const tokens = getTokenDefinitionsForActionType(lastActionType, options);
-
-	return tokens
-		.map(token => {
-			const label = TOKEN_LABELS[token.name] || token.name;
-			return `- ${label}: {{${workflowId}.${token.name}}}`;
-		})
-		.join('\n');
-}
+export function getAvailableTokensForAction(
+	workflow: WorkflowConfig,
+	actionIndex: number
+): TokenGroup[] {
+	const groups: TokenGroup[] = [];
+
+	// Workflow context tokens (always available)
+	groups.push({
+		name: 'Workflow Context',
+		tokens: WORKFLOW_CONTEXT_TOKENS
+	});
+
+	// Each previous action becomes its own top-level group
+	const previousActions = workflow.actions.slice(0, actionIndex);
+	for (const prevAction of previousActions) {
+		const options: TokenDefinitionOptions = {};
+		if (prevAction.type === 'transcription') {
+			options.timestampGranularity = prevAction.timestampGranularity;
+		}
+		const tokens = getActionTokens(prevAction.id, prevAction.type, options);
+		groups.push({
+			name: prevAction.name || `Action ${prevAction.id}`,
+			tokens
+		});
+	}
 
+	return groups;
+}
diff --git a/styles.css b/styles.css
index b4df258..8905461 100644
--- a/styles.css
+++ b/styles.css
@@ -416,35 +416,28 @@ button.mod-success:hover {
 	border-color: rgba(100, 200, 100, 0.15);
 }
 
-.available-tokens-header {
-	display: flex;
-	align-items: center;
-	gap: 8px;
-	padding: 10px 12px;
+/* Header when used with collapsible section (Setting element) */
+.available-tokens-header.setting-item {
+	padding: 8px 12px;
+	border: none;
 	cursor: pointer;
 	user-select: none;
 	background-color: rgba(100, 200, 100, 0.05);
 }
 
-.theme-dark .available-tokens-header {
+.theme-dark .available-tokens-header.setting-item {
 	background-color: rgba(100, 200, 100, 0.03);
 }
 
-.available-tokens-header:hover {
+.available-tokens-header.setting-item:hover {
 	background-color: rgba(100, 200, 100, 0.12);
 }
 
-.theme-dark .available-tokens-header:hover {
+.theme-dark .available-tokens-header.setting-item:hover {
 	background-color: rgba(100, 200, 100, 0.08);
 }
 
-.available-tokens-arrow {
-	font-size: 11px;
-	color: var(--text-faint);
-	opacity: 0.7;
-}
-
-.available-tokens-title {
+.available-tokens-header .setting-item-name {
 	font-weight: 400;
 	font-size: 13px;
 	color: var(--text-muted);
diff --git a/versions.json b/versions.json
index 5a8cf20..ba0fe2a 100644
--- a/versions.json
+++ b/versions.json
@@ -1,4 +1,5 @@
 {
 	"1.0.0": "0.15.0",
-	"1.0.3": "0.15.0"
+	"1.0.3": "0.15.0",
+	"1.0.4": "0.15.0"
 }