The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute standard git commands, including git diff and git log, which are necessary for analyzing code changes in the user's local repository.
[PROMPT_INJECTION]: The skill demonstrates a surface for indirect prompt injection as it processes untrusted data from external sources and interpolates it into context for analysis.
Ingestion points: The skill ingests untrusted data from pull request descriptions, branch names, and user-provided AI review summaries in SKILL.md (Step 1 and Step 2).
Boundary markers: There are no explicit boundary markers or instructions defined to prevent the agent from following malicious instructions potentially embedded in the ingested PR content.
Capability inventory: The skill has access to Bash, Read, Grep, and Glob tools, which could be leveraged if an injection is successful.
Sanitization: The skill lacks explicit sanitization or validation of external content before it is processed or passed to the code-reviewer agent.

reviewer-check