Agent Skills — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Agent Skills (Plugin) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
A plugin marketplace for Claude Code that adds specialized workflows and domain knowledge.
| 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. |
/plugin marketplace add skizha/agent-skills
/plugin install skizha-skills@skizha-skillsAll skills are available in your next Claude Code session under the skizha-skills namespace.
Load the repo directly as a plugin directory:
claude --plugin-dir /path/to/agent-skillsSkills 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:
Prompt Creator — activates when you say things like:
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:
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:
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:
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).
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.
To add a new skill:
skills/ with a SKILL.md fileskills/ and no changes to marketplace.json are neededMIT
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.