adcp-creative

Pass

Audited by Gen Agent Trust Hub on Feb 22, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [No Code] (SAFE): The skill consists solely of documentation and task definitions for use with MCP tools. There are no executable scripts or commands provided.
  • [Indirect Prompt Injection] (SAFE): Evaluation of ingestion surface: 1. Ingestion points: 'message' and 'assets' fields in build_creative and preview_creative tasks. 2. Boundary markers: None explicitly defined in the manifest schemas. 3. Capability inventory: Relies on external 'creative agents' via agent_url. No local subprocess, shell, or file-write capabilities are defined. 4. Sanitization: Not specified in protocol documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 22, 2026, 04:37 AM