parallel-agents
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE] (SAFE): No malicious behavior, obfuscation, or data exfiltration patterns were detected. The skill is entirely instructional and does not bundle any scripts or dependencies.
- [Indirect Prompt Injection] (SAFE): The skill specifies patterns for reading codebase content, creating an attack surface for indirect prompt injection. However, because the allowed tools (Read, Glob, Grep) are strictly read-only and no execution tools are present, the risk is negligible. Evidence: 1. Ingestion points: Local files read via 'Read' and 'Grep' tools; 2. Boundary markers: None identified in templates; 3. Capability inventory: Read-only codebase tools; 4. Sanitization: Not applicable for read-only access.
Audit Metadata