convert-prompt

Pass

Audited by Gen Agent Trust Hub on Jun 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected in the skill files.
  • [PROMPT_INJECTION]: The skill processes user-supplied data (source prompts) as its primary function, which represents an indirect injection surface. The skill includes robust procedural mitigations for this surface:
  • Ingestion points: User-provided design prompts are ingested and parsed within the 'Conversion Workflow' in SKILL.md.
  • Boundary markers: Step 14 of the workflow establishes a 'Final Prompt Audit' that checks for workspace context leakage and ensures the output is self-contained.
  • Capability inventory: The skill instructions do not request or utilize dangerous capabilities such as shell execution, file system modifications, or network operations.
  • Sanitization: The workflow requires an internal provenance ledger to remove 'unbacked additions' (Step 4) and mandates the removal of build-stack mechanics (Step 6), ensuring that the output remains faithful to the original intent without introducing unverified instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 29, 2026, 08:19 PM
Security Audit — agent-trust-hub — convert-prompt