The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute standard developer commands using git and gh (GitHub CLI) to retrieve the repository's default branch name and the diff of the current branch. These are legitimate uses of tools to gather context for a code review task.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest and process untrusted data in the form of code changes and file contents. This creates a surface for indirect prompt injection where an attacker could place malicious instructions inside a code comment or string to influence the agent's audit findings or subsequent behavior. * Ingestion points: git diff output and individual file reads during auditing. * Boundary markers: None specified to separate code content from instructions. * Capability inventory: git, gh, and the ability to read local files. * Sanitization: None mentioned for the ingested code data.

find-bugs