Version 1.10.0#23
Merged
Merged
Conversation
detect() used existsSync which returns true for files — mkdir then fails with ENOTDIR when trying to create subdirectories inside it. Co-Authored-By: Claude Opus 4.6 (1M context) via AIContext
"(optional)" label caused the AI to skip local runs every time, treating CI as the verification loop instead. Co-Authored-By: Claude via AIContext
Old rule "each new task means a new version" caused the AI to increment versions on its own instead of using the branch name. Co-Authored-By: Claude via AIContext
Co-Authored-By: Claude via AIContext
"No — create one" and "No — skip" shared a prefix that caused the AI to collapse them into a single option. Co-Authored-By: Claude via AIContext
TDD rules and planning guidelines were loaded at session start but forgotten by plan-creation time. Moving them to a dedicated prompt that loads when needed fixes the attention gap. TDD is now config-driven (tdd: true|false|ask). Co-Authored-By: Claude via AIContext
Prompts relied on stale session-start context for branch identity. Now resolve-task-naming, finish-task, and draft-pr explicitly run git status first. Also fixes git log triple-dot to double-dot in draft-pr (triple-dot includes noise from base branch commits). Co-Authored-By: Claude via AIContext
After step creation, only /run-step is relevant — /do-it is for when no step exists yet. Co-Authored-By: Claude via AIContext
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Run ID: ⛔ Files ignored due to path filters (2)
📒 Files selected for processing (3)
✅ Files skipped from review due to trivial changes (3)
📝 WalkthroughWalkthroughThis PR extracts TDD-aware planning rules into ChangesPlanning Rules Centralization & Prompt + Docs Updates
Assistant Detection, Installer, and CLI Messaging
Sequence DiagramsequenceDiagram
participant User
participant Config as Config/Validation
participant Prompt as Planning Prompt
participant Exec as CLI/Executor
User->>Config: Load repo config (includes `tdd`)
Config->>Config: Validate required keys (tdd, after_step.*, after_task.*)
alt tdd missing or ask
Config->>User: Prompt to resolve `tdd` (yes/no)
end
User->>Prompt: Invoke planning (start-feature/create-task/add-step)
Prompt->>Prompt: Read `tdd` value
alt tdd == true
Prompt->>Prompt: Emit test-first plan steps (test → implement → verify)
else tdd == false
Prompt->>Prompt: Emit standard ordered plan steps
end
Prompt->>Exec: Execute actions (run-step, finish-task, push)
Exec->>Exec: Run git checks (`git status`) and assistant detection (`isDirectory(...)`)
Exec->>User: Ask for confirmations when branch/HEAD invalid or push needed
Estimated code review effort🎯 4 (Complex) | ⏱️ ~50 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 3 | ❌ 2❌ Failed checks (1 warning, 1 inconclusive)
✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Review rate limit: 4/5 reviews remaining, refill in 12 minutes. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 4
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
bin/aicontext.js (1)
58-76:⚠️ Potential issue | 🟠 Major | ⚡ Quick winHandle assistant-path file collisions before install (regression in
add-assistant).Directory-only detection is correct, but on Line 1059/Line 1068 a file at
.claude/.cursor/.codexis now treated as “not installed,” then install crashes when writing directories. Add an explicit preflight conflict check inaddAssistant(and ideally shared for install flows).💡 Suggested fix
async function addAssistant(name, targetDir, skipConfirm = false) { const target = path.resolve(targetDir || '.'); @@ const assistant = ASSISTANTS[name]; + const assistantPath = path.join(target, assistant.folder.replace(/\/$/, '')); + const isDirAssistant = ['claude', 'cursor', 'codex'].includes(name); + if (isDirAssistant && fs.existsSync(assistantPath) && !isDirectory(assistantPath)) { + log(`Cannot install ${assistant.label}: ${assistantPath} exists as a file.`, 'red'); + log('Rename/remove the file, then retry.', 'dim'); + return; + } + if (assistant.detect(target)) { log(`${assistant.label} (${assistant.folder}) is already installed.`, 'yellow'); log(`Run \`aicontext update\` to refresh it.`, 'dim'); return; }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@bin/aicontext.js` around lines 58 - 76, The install path check is missing a preflight for file-vs-directory collisions causing installs to crash when a file exists at the assistant folder (e.g. ".claude", ".cursor", ".codex"); update addAssistant to call a shared helper (e.g. ensureAssistantTargetIsDirectory or validateInstallTarget) before performing installs so it checks path existence and type and fails fast or prompts when the target exists as a file, and reuse that helper from install flows that call copyRecursive, copyFrameworkAgents, and copyFrameworkSkills; also ensure the per-assistant detect methods remain directory-only but the installer always verifies the target is a directory (or empty/non-existent) before writing.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In @.aicontext/prompts/add-step.md:
- Line 36: Update the confirmation text that currently reads "Show the step that
was added" to handle both the new-add and existing-step flows: detect the
"existing-step" path and use a neutral message like "Show the step that was
added or the existing step referenced" (or similar) so it does not assume a new
add; also prompt the user to execute the step with "/run-step". Locate the
phrase "Show the step that was added" and the "existing-step" path in
add-step.md and replace the hardcoded confirmation with this conditional/neutral
wording and include the suggested "/run-step" instruction.
In @.aicontext/prompts/finish-task.md:
- Line 68: The prompt's instruction for after_task.push that runs "git status"
then "git push -u origin {current-branch}" lacks explicit abort checks; update
the prompt so after_task.push first inspects git status output and aborts if any
fail-fast conditions are detected (detached HEAD, no current branch name,
unmerged/uncommitted changes, or branch mismatch against expected
{current-branch}), instructing the agent to stop and report the exact condition
instead of running git push; reference the after_task.push step and the git push
-u origin {current-branch} action so the agent performs these verifications and
only proceeds to push when all checks pass.
In @.aicontext/prompts/resolve-task-naming.md:
- Line 40: The fallback description incorrectly says to "extract branch name via
task_naming.branch_pattern" — update the wording so it clearly states that when
task_naming.source is unset you should infer package-json if package.json has a
version field, otherwise use git-branch and run git status to extract the
{version} token using task_naming.branch_pattern (i.e., the pattern is used to
derive {version}, not the branch name); adjust the sentence to replace "branch
name" with "{version}" and make clear the pattern maps branch text to the
{version} value.
In @.aicontext/prompts/review-task.md:
- Line 11: The TDD enforcement in the review-task prompt currently flags
test-first violations whenever config.tdd is true; update the check so it
mirrors plan-steps.md by skipping the TDD gate when the plan indicates
"no-tests" (i.e., when the task/plan contains the no-tests marker or explicit
absence of tests). Concretely, modify the TDD structure logic that reads the tdd
flag to also inspect the plan metadata (the same no-tests condition used in
plan-steps.md) and only enforce test-first rules for functions/steps under
/review-task when tdd is true AND no-tests is not set.
---
Outside diff comments:
In `@bin/aicontext.js`:
- Around line 58-76: The install path check is missing a preflight for
file-vs-directory collisions causing installs to crash when a file exists at the
assistant folder (e.g. ".claude", ".cursor", ".codex"); update addAssistant to
call a shared helper (e.g. ensureAssistantTargetIsDirectory or
validateInstallTarget) before performing installs so it checks path existence
and type and fails fast or prompts when the target exists as a file, and reuse
that helper from install flows that call copyRecursive, copyFrameworkAgents, and
copyFrameworkSkills; also ensure the per-assistant detect methods remain
directory-only but the installer always verifies the target is a directory (or
empty/non-existent) before writing.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
Run ID: e5a6b109-8206-41c6-a7ed-0cde103d8888
📒 Files selected for processing (19)
.aicontext/prompts/add-step.md.aicontext/prompts/aic-help.md.aicontext/prompts/create-task.md.aicontext/prompts/draft-pr.md.aicontext/prompts/ensure-config.md.aicontext/prompts/finish-task.md.aicontext/prompts/gh-fix-tests.md.aicontext/prompts/plan-steps.md.aicontext/prompts/plan-tasks.md.aicontext/prompts/resolve-asks.md.aicontext/prompts/resolve-task-naming.md.aicontext/prompts/review-task.md.aicontext/prompts/start-feature.md.aicontext/prompts/start.md.aicontext/readme.md.aicontext/rules/process.md.aicontext/rules/standards.md.aicontext/templates/config.template.ymlbin/aicontext.js
…ts gate
Adds fail-fast before push when HEAD is detached or branch name is
unavailable. Fixes "extract branch name" → "extract {version}" wording.
Mirrors plan-steps.md "skip when no tests" rule in review-task TDD check.
Co-Authored-By: Claude via AIContext
Co-Authored-By: Claude via AIContext
Configurable TDD, plan-steps extraction, branch verification, git range fix, misleading next-action fix. Co-Authored-By: Claude via AIContext
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
tdd: true | false | ask) — agents assess testability per step and apply test-first only when configuredplan-steps.mdprompt loaded at plan-creation time, reducing session-start context by ~62 lines fromprocess.mdgit statusbefore remote operations (push, PR creation, version extraction from branch)git logtriple-dot to double-dot indraft-pr.mdfor correct commit range/do-itsuggestion after/add-stepwhen step already existsTest plan
node --test test/*.test.jspasses (171 tests)/create-taskon agit-branchsource project runsgit statusand extracts version correctly/finish-taskwithpush: trueverifies branch before pushing/add-steponly suggests/run-stepin confirmationtdd: trueproduces test-first plan steps;tdd: falseskips TDD assessmentSummary by CodeRabbit
New Features
Improvements
Documentation