The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes standard system commands such as find, git log, and npx skills list to inventory project files and identify installed skills. These operations are restricted to information gathering and discovery within the local environment.
[DATA_EXPOSURE]: The skill skims repository documentation (e.g., README.md, VISION.md) and git history to generate project summaries. This data analysis occurs entirely within the agent's context and is not transmitted to external services.
[INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted repository content, creating a potential surface for indirect prompt injection.
Ingestion points: The skill reads README.md, AGENTS.md, git log, and various documentation files in the docs/ directory.
Boundary markers: No specific delimiters are used when processing file content, although the agent's output is structured.
Capability inventory: This skill is explicitly read-only; while it recommends other tools with write capabilities (like ship-it or merge-it), it does not execute them autonomously.
Sanitization: The skill does not perform sanitization on the ingested repository data.

idk-now