lean-canvas
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill includes an attack surface for indirect prompt injection as it ingests data from local files and external sources via the SandBase API (search/scraping). However, the risk is addressed through instructional guardrails that direct the agent to preserve authorization boundaries, keep core work local, and minimize data transmission.
- [SAFE]: The skill follows security best practices by requiring user confirmation for costs and unit prices before execution and by prioritizing local tools and data sources over external ones.
Audit Metadata