The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest information from external code, configurations, and comments to generate documentation. This represents an attack surface where malicious instructions hidden in code comments or docstrings could influence the agent's behavior during the documentation generation process.
Ingestion points: Instructions in SKILL.md direct the agent to "Extract information from code, configs, and comments."
Boundary markers: None identified in the prompt templates or instructions to warn the agent about ignoring instructions embedded in the processed data.
Capability inventory: The skill uses tools to read files, write files, and potentially execute scripts for documentation generation (e.g., sphinx-build, redocly).
Sanitization: There are no explicit steps for sanitizing or escaping the content extracted from the code before it is used in documentation templates.
[COMMAND_EXECUTION]: The resources/implementation-playbook.md provides examples of shell commands for CI/CD pipelines (e.g., pip install, npm install, sphinx-build). While these are intended as documentation templates for the user, an agent might inadvertently attempt to execute these if the user prompt is ambiguous.
[SAFE]: External references to Swagger UI (via cdn.jsdelivr.net) and standard GitHub Actions (actions/checkout, actions/setup-python, peaceiris/actions-gh-pages) are recognized as well-known and trusted services/registries.

code-documentation-doc-generate