agent-md-creator
Pass
Audited by Gen Agent Trust Hub on Apr 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a documentation tool that assists in drafting project-specific instructions for agents based on verified repository facts.
- [EXTERNAL_DOWNLOADS]: The skill references trusted external documentation, such as the GitHub Blog and official agents.md specifications, to provide grounding for its guidance.
- [COMMAND_EXECUTION]: While the skill identifies and records development commands (e.g., pytest, npm run lint) for documentation purposes, it does not execute these commands itself in a malicious or unauthorized manner.
- [INDIRECT_PROMPT_INJECTION]: The skill includes an ingestion surface for processing untrusted repository data but follows structural best practices to mitigate risks.
- Ingestion points: Reads existing repository files such as AGENTS.md, README.md, CI workflows, and tool-specific instruction files.
- Boundary markers: Recommends establishing AGENTS.md as a canonical source with clear headers and scoped instructions.
- Capability inventory: Documents commands for testing, linting, and debugging that the target agent is expected to utilize.
- Sanitization: Relies on the AI agent's synthesis process to transform ingested file content into a technical markdown format.
Audit Metadata