Skip to content

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
 
 
 
 
 
 
 
 

Repository files navigation

agent-skills

A plugin marketplace for Claude Code that adds specialized workflows and domain knowledge.

Available skills

Skill What it does
humanize Detects and rewrites 24 categories of AI-generated text patterns. Based on Wikipedia's "Signs of AI writing" guide.
prompt-creator Expert prompt engineering for AI models (especially Claude). Creates prompts from scratch, improves underperforming prompts, designs system prompts, structures chain-of-thought and few-shot examples, and builds RAG and agentic prompts. Based on Anthropic's official prompt engineering docs.
sql-server-query-tuner SQL Server query performance tuning — analyzes execution plans, identifies anti-patterns, recommends indexes, rewrites slow queries, and runs DMV diagnostics against live instances.
linkedin-articles Write and optimize LinkedIn articles — 5 structure templates, 8 hook formulas, headline patterns, SEO keyword placement, and hashtag strategy.
excalidraw-diagram Generate beautiful .excalidraw diagram files for architectures, workflows, and concepts — with visual argument design, evidence artifacts, and a Playwright-based render pipeline.

Installation

From the marketplace

/plugin marketplace add skizha/agent-skills
/plugin install skizha-skills@skizha-skills

All skills are available in your next Claude Code session under the skizha-skills namespace.

Local testing

Load the repo directly as a plugin directory:

claude --plugin-dir /path/to/agent-skills

Usage

Skills trigger automatically based on your prompt. You can also invoke them directly:

Skill Direct invocation
humanize /skizha-skills:humanize
prompt-creator /skizha-skills:prompt-creator
sql-server-query-tuner /skizha-skills:sql-server-query-tuner
linkedin-articles /skizha-skills:linkedin-articles
excalidraw-diagram /skizha-skills:excalidraw-diagram

Humanize — activates when you say things like:

  • "humanize this text"
  • "make this sound less robotic"
  • "remove AI writing patterns from this file"

Prompt Creator — activates when you say things like:

  • "write a prompt for a customer service chatbot"
  • "create a system prompt that outputs JSON"
  • "improve this prompt — it keeps hallucinating"
  • "help me prompt engineer a RAG pipeline"
  • "why isn't my prompt working?"
  • "make a few-shot prompt for sentiment classification"

Covers all major techniques: role/persona, few-shot, chain-of-thought, prompt chaining, structured output, RAG, agentic prompts, and meta-prompting. Includes Claude-specific best practices (XML tags, extended thinking, long context ordering) and a diagnostics guide for fixing underperforming prompts. Based on official docs from Anthropic, Microsoft Azure OpenAI, and OpenAI.

SQL Server Query Tuner — activates when you say things like:

  • "tune this query"
  • "why is this SQL slow"
  • "check my database for missing indexes"
  • "analyze this execution plan"
  • "optimize this SQL Server query"

Supports both pasting queries/execution plans for analysis and running DMV diagnostics against a live SQL Server instance.

LinkedIn Articles — activates when you say things like:

  • "write a LinkedIn article about [topic]"
  • "optimize my LinkedIn article"
  • "write a hook for my article"
  • "add hashtags to this article"
  • "improve my article structure"

Includes 5 article structure templates (Problem-Solution, Listicle, Story-Driven, How-To, Contrarian), 8 hook formulas with examples, headline patterns, SEO keyword placement guidance, and a hashtag strategy (broad/mid/niche mix). Delivers ready-to-paste output with hashtags.

Excalidraw Diagram — activates when you say things like:

  • "create a diagram for my authentication flow"
  • "draw an architecture diagram for this system"
  • "visualize this workflow as an Excalidraw diagram"
  • "make an Excalidraw diagram showing the data pipeline"

Generates .excalidraw JSON files following visual argument design principles — fan-outs for one-to-many, timelines for sequences, convergence for aggregation. Includes a Playwright-based render pipeline to preview diagrams as PNG and iteratively fix layout issues. Customizable color palette and a library of element templates. Requires Python + uv for the optional render pipeline (one-time setup).

Marketplace structure

agent-skills/
├── .claude-plugin/
│   ├── plugin.json         # plugin manifest
│   └── marketplace.json    # marketplace manifest
└── skills/
    └── skill-name/
        ├── SKILL.md           # required -- instructions and metadata
        ├── references/        # optional -- detailed docs loaded on demand
        ├── scripts/           # optional -- executable code
        └── assets/            # optional -- templates, images, etc.

The marketplace.json defines a single plugin (skizha-skills) that bundles all skills. Each skill lives under skills/ with at least a SKILL.md containing YAML frontmatter (description field) and markdown instructions. Skills are auto-discovered from the skills/ directory.

Contributing

To add a new skill:

  1. Create a folder under skills/ with a SKILL.md file
  2. Submit a PR — the skill is auto-discovered from skills/ and no changes to marketplace.json are needed

License

MIT

About

A collection of Claude Code skills -- reusable plugins that extend Claude with specialized workflows and domain knowledge.

Topics

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages