grill-me-flow

Pass

Audited by Gen Agent Trust Hub on Jul 20, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a workflow coordinator that gathers user requirements before proceeding to planning. No security issues were detected.
  • [PROMPT_INJECTION]: The instructions provide strict operational boundaries (gates) to prevent the agent from acting prematurely, which enhances safety. No adversarial prompt injection patterns were found.
  • [DATA_EXFILTRATION]: All file operations are restricted to the local .agents/temp/ workspace. No network calls, credential access, or data exfiltration patterns were identified.
  • [REMOTE_CODE_EXECUTION]: The skill does not include any external dependencies, package installations, or remote script downloads. It only interacts with internal skill flow references.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 20, 2026, 01:04 PM
Security Audit — agent-trust-hub — grill-me-flow