skills/elihuvillaraus/skills/ralph/Gen Agent Trust Hub

ralph

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via the PRD files it processes.
  • Ingestion points: The agent loads a PRD file based on user input (Step 1 in SKILL.md).
  • Boundary markers: There are no explicit delimiters or specific instructions to the agent to disregard natural language commands or malicious instructions embedded within the PRD file content.
  • Capability inventory: The skill is capable of performing filesystem modifications to codebase files (Step 4) and executing shell commands to run 'Quality Gates' defined in the document headers (Step 5).
  • Sanitization: No sanitization, escaping, or validation of the PRD content or the 'Quality Gate' commands is mentioned or implemented before the agent acts on them.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 02:51 PM
Security Audit — agent-trust-hub — ralph