The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external data from local codebases (source code files and markdown documentation) to verify alignment. While this creates an inherent attack surface for indirect prompt injection, it is central to the skill's primary function and no malicious patterns were detected.
Ingestion points: Processes files discovered via Glob and Grep tools, specifically parsing docstrings, type hints, and markdown files as described in Phase 1 and 2 of the reference guide.
Boundary markers: The instructions do not explicitly define boundary markers for separated content, though they recommend structured extraction.
Capability inventory: The skill utilizes Read, Grep, Glob, and TodoWrite tools, providing the agent with the ability to read project files and write to them (specifically for fix generation in Phase 5).
Sanitization: No specific sanitization or filtering of external content is mentioned before the data is processed by the agent.
[COMMAND_EXECUTION]: The skill documentation mentions several CLI tools for integration (e.g., ts-node, typedoc, vitest, pytest, mypy). These are referenced as standard development practices for CI/CD integration and do not involve executing untrusted strings from external sources.

documentation-alignment