The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from repository files and PR diffs which may contain adversarial instructions designed to bypass review conventions or force a specific verdict.\n
Ingestion points: Uses git diff and gh pr diff to read change sets and file names in SKILL.md.\n
Boundary markers: No explicit delimiters or safety instructions are defined to separate untrusted file content from the agent's orchestration logic.\n
Capability inventory: Includes the ability to commit changes via git, post reviews via gh api, and execute a local Python script conventions.py.\n
Sanitization: The instructions do not specify any validation or sanitization steps for the data retrieved from git or GitHub before it is processed by the agent or sub-skills.

quality-gate