agents-md
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and analyze untrusted project data (such as package manager lock files, linter configs, and CI scripts) to generate the
AGENTS.mdfile. This creates an attack surface where maliciously crafted project metadata could influence the agent's behavior during the analysis or through the content of the generated documentation.\n - Ingestion points: The skill explicitly directs the agent to 'Analyze the project' by checking lock files (
pnpm-lock.yaml,yarn.lock, etc.), linter/formatter configs (.eslintrc,ruff.toml), and CI/build commands (Makefile,package.json) in the project root.\n - Boundary markers: There are no instructions providing delimiters or warnings to the agent to disregard instructions found within these ingested files.\n
- Capability inventory: The skill leverages the agent's ability to read repository files, write documentation files (
AGENTS.md), and create symlinks (ln -s).\n - Sanitization: The skill does not specify any sanitization, filtering, or validation steps for the content extracted from the repository files before it is processed or written into the final documentation.
Audit Metadata