diagnose
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill's primary logic is focused on business diagnosis and strategic analysis. No malicious patterns such as credential harvesting, unauthorized network connections, or persistence mechanisms were detected. The included support script scripts/bootstrap-experience.ts implements safety checks to prevent directory traversal via symlinks.- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from web searches and user-provided evidence (e.g., CSVs, reports), which constitutes a potential surface for indirect prompt injection. This is a characteristic of the skill's intended research functionality.
- Ingestion points: external-check-agent.md (WebSearch results) and verdict-agent.md (User-provided evidence).
- Boundary markers: Prompt instructions do not explicitly implement markers to delimit untrusted content.
- Capability inventory: The skill uses tools including Bash, WebSearch, and WebFetch across its agents.
- Sanitization: No automated sanitization of external content is specified in the provided files.
Audit Metadata