The Agent Skills Directory

[SAFE]: No malicious patterns or unauthorized behaviors were detected. The skill's use of the Bash tool to execute git, npm test, and npm run lint is typical for automated code auditing and development workflows.
[PROMPT_INJECTION]: The skill exhibits an inherent attack surface for indirect prompt injection because it ingests and processes untrusted code changes from external repositories.
Ingestion points: Code changes identified via git diff and content accessed through the Read tool in SKILL.md.
Boundary markers: Absent; the instructions do not explicitly tell the agent to ignore instructions embedded within the code being reviewed.
Capability inventory: Includes the Bash tool, which is used to run project-level scripts like npm test and npm run lint in SKILL.md.
Sanitization: None; the skill evaluates implementation details in their raw format to provide accurate feedback.

code-reviewer