code-review

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it ingests and processes untrusted code changes from pull requests and merge requests.
  • Ingestion points: The agent reads the diff, surrounding files, and project structure of external code changes (README.md, SKILL.md).
  • Boundary markers: There are no explicit delimiters or instructions to treat the reviewed code as untrusted data or to ignore embedded instructions within that code.
  • Capability inventory: While the instructions specify providing textual feedback and not modifying code, the underlying agent may possess broader capabilities (file system access, shell execution) that could be targeted by an injection.
  • Sanitization: No validation or sanitization is performed on the content of the code being reviewed to prevent malicious instructions in comments or strings from influencing the agent's behavior.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 01:58 PM
Security Audit — agent-trust-hub — code-review