- Home
- Skills
- Code Quality & Review
- Skill Readiness Auditor
Skill Readiness Auditor
Local pre-publish auditor for agent skills that validates metadata, safety signals, and documentation readiness.
$14.99
Works with the AI Agents you already use
Skill Readiness Auditor
Example session with this skill installed
Audit the local directory 'my-new-skill' and fail if any high-level issues are found. Emit a JSON report.
- Read your context and instructions
- Compiled the readiness auditor
Audit completed for 1 target. The readiness check identified missing documentation and placeholder text.
Audit Report: my-new-skill
| Category | Level | Rule | File | Hint |
|---|---|---|---|---|
| Readiness | High | ready_placeholder_residue | SKILL.md | Remove "TODO: add real instructions" before release. |
| Safety | Medium | safety_shell_execution | scripts/run.py | AST found subprocess.run with shell=True. Use a list of args instead. |
| Readiness | Low | ready_readme_missing | README.md | Add a root README.md to describe installation and setup. |
| Readiness | Low | ready_examples_missing | SKILL.md | Add an 'Example' section with sample inputs and outputs. |
Summary
- Status:
blocked - Level:
high - Violations: 1 high, 1 medium, 2 low
- Exit Code: 1
Next steps
- Remove the
TODOcomments fromSKILL.mdto pass the placeholder check. - Refactor
scripts/run.pyto avoid raw shell execution. - Re-run with
--require-readmeto ensure your documentation meets baseline standards.
Connects securely to your tools. The creator never sees your data.
About this skill
Publishing an AI agent skill with broken frontmatter, missing caveats, or risky code patterns leads to rejection or poor user experience. The Skill Readiness Auditor provides a local, offline gate to ensure your package meets professional standards before you push to a registry.
What it does
- Validation checks SKILL.md frontmatter, trigger specificity, and structural completeness.
- Safety Scanning uses AST parsing on Python scripts to detect exfiltration, execution, or credential-access signals without running the code.
- Wording Analysis flags overclaims like "100% compatible" or "guaranteed safe" that violate marketplace policies.
- Consistency Checks verifies that claimed scripts actually exist and that license fields match root LICENSE files.
- CI/CD Integration provides machine-readable JSON reports and configurable exit codes for automated release pipelines.
How it works
- Target selection: Point the tool at a local skill directory or a specific SKILL.md file.
- Static analysis: Run the auditor via the CLI to parse prose and code using built-in heuristic rules.
- Report generation: Review the human-readable summary or structured JSON findings categorized by level: safe to critical.
- Fixing findings: Use the provided hints to resolve placeholder residue, missing examples, or risky script patterns.
Frameworks & tools
This tool is a zero-dependency CLI compatible with Python 3.9+. It uses the Python standard library ast module for static code analysis and does not require third-party packages or network access.
Why this beats prompting it yourself
Manual reviews often miss small details like frontmatter syntax errors or subtle "download-to-shell" chains in scripts. This tool provides a deterministic, repeatable checklist that doesn't suffer from LLM hallucination or context window limits when scanning large directories.
Use cases
- Audit a local skill directory before publishing to Agensi.
- Gate a GitHub Action to prevent merging skills with missing documentation.
- Batch scan a folder of agent tools to find inconsistent licensing or metadata.
- Sanitize third-party skills for safety signals before adding them to a local library.
Known limitations
Heuristic matching can flag benign wording or miss obfuscated code. It is a pre-check tool, not a substitute for formal security audits or legal review.
Heuristic signals may produce false positives on benign wording. AST parsing is limited to Python; other languages use text matching. Requires manual review of high/critical findings.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 7 days ago
Requires