ralph

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted PRD documents and converts them into a structured format for autonomous execution, creating an indirect injection surface.
  • Ingestion points: External markdown or text PRD files processed by the agent as described in 'The Job' section.
  • Boundary markers: The instructions do not define delimiters or safety markers to isolate processed data from agent instructions within the requirements.
  • Capability inventory: The skill utilizes 'Read' and 'Write' tools to manipulate system files, including 'prd.json' and the 'archive' directory.
  • Sanitization: No validation or sanitization mechanisms are in place to inspect the contents of the input PRD before inclusion in the final JSON output.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 02:27 AM