to-prd

Pass

Audited by Gen Agent Trust Hub on Jun 26, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security risks were identified in the skill's instructions or metadata. The skill focuses on standard project management documentation tasks.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it ingests untrusted data from the local repository and the current conversation context to generate its output.
  • Ingestion points: Processes codebase content and conversation history as specified in the 'Process' section of SKILL.md.
  • Boundary markers: No specific boundary markers or 'ignore' instructions are used for the input data, though the output is structured via a <prd-template>.
  • Capability inventory: Performs repository exploration and publishes to an external issue tracker.
  • Sanitization: No explicit sanitization of the synthesized content is mentioned before publication.
  • [DATA_EXFILTRATION]: The skill is designed to send synthesized information to an external 'project issue tracker'. While this involves network transmission of project-related data, it is the explicitly stated primary purpose of the skill and relies on the agent's pre-configured tool permissions (e.g., GitHub CLI).
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 26, 2026, 02:40 PM
Security Audit — agent-trust-hub — to-prd