The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various shell commands to facilitate its workflow, including repository state detection (git rev-parse), extracting code changes (git diff), and stack identification (cat package.json | jq). It also performs repository modifications through git add, git commit, and git push if changes are applied.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) because it ingests and processes untrusted data in the form of git diffs from external branches. Malicious instructions embedded within the code being reviewed (e.g., in comments or string literals) could potentially influence the behavior of the sub-agents or the logic used to apply auto-fixes.
Ingestion points: Git diff output and file contents read during the review process in SKILL.md.
Boundary markers: The skill lacks formal delimiters or "ignore instructions" warnings for the diff data passed to reviewers.
Capability inventory: The lead agent has the capability to modify the codebase via the Edit tool and commit changes via git as described in SKILL.md.
Sanitization: No explicit sanitization or filtering of the diff content is performed before it is passed to the LLM for review.
[NO_CODE]: This is a purely instructional skill defined within SKILL.md. It does not include external scripts, binaries, or configuration files, relying instead on the agent's ability to use standard tools and other available skills.

quality-gate