skill-pressure-test
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions define a benign analytical role focused on identifying flaws in user-provided designs. The agent is directed to use repository data (code, config, history) to verify facts, which is standard developer tool behavior. No indicators of prompt injection, obfuscation, or unauthorized network activity were found.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes untrusted user plans and potential external sources.
- Ingestion points: Ingests plans, specifications, and external sources as defined in the 'Workflow' and 'Provider Or Data Priority' sections of SKILL.md.
- Boundary markers: No specific delimiters or instruction-override protections are defined for external data processed by the agent.
- Capability inventory: The agent can read the local filesystem, git history, config files, and execute unspecified tools as defined in the 'Workflow' section.
- Sanitization: No explicit sanitization or filtering of external content is mentioned.
Audit Metadata