afk-code-review-verdicts

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes findings from an automated code review tool (afk-code-review). This creates an attack surface where malicious instructions embedded in the code under review could be passed to the agent via the review tool's output.
  • Ingestion points: Processes the complete output from the afk-code-review skill (SKILL.md).
  • Boundary markers: The skill instructs the agent to present output verbatim and append a specific header (## Verified verdicts), which helps structure the context but does not fully isolate the agent from instructions within the ingested data.
  • Capability inventory: The agent has permission to read local source code and provide suggested code fixes.
  • Sanitization: The instructions mitigate risk by requiring the agent to treat input as "unverified," inspect findings against the "actual code," and strictly prohibits automatic code changes without prior discussion.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 01:32 AM
Security Audit — agent-trust-hub — afk-code-review-verdicts