cy-create-prd
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill's primary functions—searching the codebase, performing web research, and writing documentation to the
.compozy/tasks/directory—are consistent with its stated purpose of creating PRDs. These operations are performed using standard agent capabilities and do not exhibit malicious patterns such as credential theft, persistence, or unauthorized privilege escalation.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the web and the local codebase to inform its requirements gathering. This inherently creates a surface for indirect prompt injection where malicious content in search results could attempt to influence agent actions.\n - Ingestion points: Data is ingested from codebase search results (Track A) and web search results (Track B) as specified in the workflow section of SKILL.md.\n
- Boundary markers: There are no specific boundary markers or instructions defined to prevent the agent from following instructions that might be embedded in the retrieved research data.\n
- Capability inventory: The skill possesses capabilities for codebase searching, web searching, and file writing (documented in steps 2, 4, and 6 of the workflow).\n
- Sanitization: No explicit sanitization or validation of the content retrieved from web or codebase searches is performed before the data is integrated into the final PRD and ADR artifacts.
Audit Metadata