convert-prd-to-json
Pass
Audited by Gen Agent Trust Hub on Apr 1, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted input from markdown files (prd.md) to generate task plans, creating a surface for indirect prompt injection where malicious content in the PRD could influence agent behavior.\n
- Ingestion points: Read access to issues/{ISSUE_NUMBER}/prd.md\n
- Boundary markers: Absent; instructions do not specify delimiters or warnings for embedded content.\n
- Capability inventory: File system write access, directory creation, and file deletion.\n
- Sanitization: Absent; content is converted directly without validation or filtering.\n- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands for archiving stale task plans and removing progress logs.\n
- Evidence: Archive logic uses mkdir, cp, and rm via the shell.\n
- Evidence: rm -f issues/{ISSUE_NUMBER}/progress.txt
Audit Metadata