The Agent Skills Directory

[SAFE]: The skill's instructions, metadata, and templates are consistent with its stated purpose of documentation management. No attempts to override safety protocols or access sensitive information were found.\n- [COMMAND_EXECUTION]: The skill executes a bundled script scripts/doc-inventory.py using python3 to audit documentation coverage. The script was analyzed and found to perform only local file analysis and regex-based symbol extraction without any network access, file-writing, or dangerous subprocess calls.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests source code from the project environment to generate documentation.\n
Ingestion points: Reads files with extensions including .py, .js, .ts, .go, .php, and .dart in the target directory.\n
Boundary markers: None identified in the skill instructions or templates to delimit extracted code content from agent instructions.\n
Capability inventory: The skill uses the Write tool to modify documentation files and python3 for running the local inventory script.\n
Sanitization: No explicit sanitization or filtering of comments or docstrings extracted from source code is mentioned.

docs