The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon external data sources, specifically implementation specifications and task tickets.
Ingestion points: The skill reads work descriptions from user-provided specs or tickets, and additionally reads project configuration from docs/agents/stack.md to determine which sub-skills to load.
Capability inventory: The skill has instructions to run typechecking, execute test suites, perform code reviews, and commit changes to the current branch via the host's tools.
Boundary markers: The instructions lack explicit delimiters or specific warnings to ignore embedded instructions within the ingested data.
Sanitization: There is no mention of sanitization or validation for the content of the specifications or tickets before processing.
[COMMAND_EXECUTION]: The skill performs standard software development tasks that involve executing local tools.
Evidence: Instructions to "Run typechecking regularly", "single test files", "full test suite", and "Commit your work to the current branch". These are routine operations for a developer-oriented agent skill.

implement