The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes git diff to identify changes within the local repository for analysis.
[DATA_EXFILTRATION]: Shares code content with the well-known Codex service via the MCP protocol for the purpose of validating review findings.
[PROMPT_INJECTION]: As the skill processes code from external files and git diffs, it is inherently subject to indirect prompt injection through malicious instructions embedded in code comments.
Ingestion points: Raw output from git diff and contents of specific files provided as scope.
Boundary markers: No explicit boundary markers or isolation instructions are present in the delegation prompt.
Capability inventory: File system read access (git diff), internal agent delegation, and external tool communication (mcp__x__ask_codex).
Sanitization: There is no evidence of pre-processing or sanitization of the input code to prevent the execution of embedded instructions.

code-review