pipeline
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWNO_CODE
Full Analysis
- [No Code] (SAFE): The skill consists entirely of markdown documentation and instructional templates without associated scripts or executable files.
- [Prompt Injection] (SAFE): The instructions describe a benign task management pattern and do not contain patterns intended to subvert agent constraints or extract system prompts.
- [Indirect Prompt Injection] (LOW): The skill creates a framework for processing task artifacts which could potentially contain untrusted data. Evidence: 1. Ingestion: Artifacts and stage outputs. 2. Boundary markers: Template structures only. 3. Capability inventory: bash and file tools. 4. Sanitization: None. Severity is low as this is an abstract structural framework rather than a functional data processor.
- [Data Exposure & Exfiltration] (SAFE): No instructions for accessing sensitive credentials, private keys, or performing unauthorized network operations were detected.
Audit Metadata