The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it processes repository files and commit history that could contain untrusted data from third-party contributions.\n
Ingestion points: Reads configuration files like AGENTS.md and SKILL.md, and processes git log output via the scripts/codex_docs_drift.py script.\n
Boundary markers: Absent. The skill does not define clear delimiters or instructions for the agent to ignore commands that might be embedded within the analyzed repository files or commit messages.\n
Capability inventory: The skill includes an executable Python script and explicitly instructs the agent to propose and apply patches to repository documentation.\n
Sanitization: No sanitization or escaping is performed on the content ingested from the repository or git logs before it is analyzed by the agent.\n- [COMMAND_EXECUTION]: The skill executes system commands via an included utility script.\n
Evidence: The script scripts/codex_docs_drift.py uses subprocess.run to execute git commands (e.g., git log, git ls-files). While the script follows security best practices by using list-based arguments to avoid shell injection, it still performs automated command execution based on the repository's state.

docs-drift