experience-review
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill requires executing local shell commands to manage the lifecycle of the system under test (e.g., starting services, running migrations) and to perform final evidence comparison using a local script:
node .claude/scripts/lib/experience-verification.cjs compare. - [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted external data to perform its review and repair tasks, creating an indirect injection surface.
- Ingestion points: The agent is instructed to read console logs, API responses, UI content, and persisted artifacts (Step 3 and 4).
- Boundary markers: No specific delimiters or instruction-ignore blocks are defined for the ingested content.
- Capability inventory: The agent can execute arbitrary shell commands and use a code-modification tool ($fix) based on its analysis of this data.
- Sanitization: The instructions recommend redacting secrets but do not specify sanitization of active instructions within the data.
Audit Metadata