verify-workflow-receiving-review
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted external data (code review feedback) which could contain malicious instructions.
- Ingestion points: External review feedback from human partners or third-party reviewers is processed in Step 1 and Step 2 of the workflow.
- Boundary markers: Absent. The skill does not define specific delimiters to isolate the untrusted feedback content within the agent's context.
- Capability inventory: The skill authorizes code modification and test execution via Step 6 and the associated 'build-workflow-execute' reference.
- Sanitization: Present. The workflow includes a mandatory 'VERIFY' step (Step 3) where the agent must confirm the feedback exists in the codebase and an 'EVALUATE' step (Step 4) to assess technical risk and adherence to existing patterns before implementation. This logical gating significantly reduces the risk of executing injected commands.
- [PROMPT_INJECTION]: The skill uses authoritative language such as 'Iron Law' and 'HARD-GATE'. These are used to reinforce the agent's internal logic and defensive posture against blind obedience to external feedback, rather than attempting to bypass the agent's own safety guidelines.
Audit Metadata