recipe-validate
Pass
Audited by Gen Agent Trust Hub on Jul 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes 'code execution' and a 'codebase-analyzer' sub-agent for feasibility validation ('code spikes'). These high-capability tools are used within a professional context and are governed by a mandatory blocking step where the user must explicitly confirm the validation design before execution.
- [EXTERNAL_DOWNLOADS]: The skill performs web searches for market research and viability analysis, which involves fetching and processing data from external, untrusted sources.
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it ingests untrusted data from web search results and user-supplied $ARGUMENTS. These inputs flow into the orchestrator and the 'hypothesis-verifier' sub-agent without explicit boundary markers or sanitization, though the risk is mitigated by structured workflow steps and mandatory human review.
Audit Metadata