metrics-dashboard
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from multiple local files provided by the user, which could contain malicious instructions designed to manipulate the agent.\n
- Ingestion points:
founder/facts.md,founder/product-brief.md,founder/mvp-scope.md,founder/pricing-strategy.md, andfounder/go-to-market.md.\n - Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the ingested data.\n
- Capability inventory: The agent has access to
WebSearch,WebFetch, and file writing (Edit) within thefounder/directory.\n - Sanitization: No sanitization, validation, or filtering is applied to the content read from these files.\n- [DATA_EXFILTRATION]: The skill facilitates a potential channel for sensitive business data exposure through the use of web-based tools.\n
- Evidence: The skill's workflow involves reading business documentation and subsequently using
WebSearchto find information on external tools. Sensitive information extracted from the local documents could inadvertently be included in search queries directed to external services.
Audit Metadata