team-composition-patterns
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE] (SAFE): Analysis of the skill files (SKILL.md, references/agent-type-selection.md, references/preset-teams.md) confirms they are strictly informational markdown. No executable scripts, binaries, or configuration files capable of executing commands were found. No package manifests or remote code execution patterns were detected.
- [Indirect Prompt Injection] (SAFE): The documentation defines workflows for 'Research' and 'Security' teams that ingest untrusted data from the web or codebase. 1. Ingestion points: WebFetch and Read tool usage mentioned in references/preset-teams.md. 2. Boundary markers: Absent in provided task templates. 3. Capability inventory: Agent roles have access to high-privilege tools including Bash, Write, and Edit according to references/agent-type-selection.md. 4. Sanitization: No sanitization or escaping guidelines are provided for the ingested content. As the skill contains no implementation code, the risk is restricted to the documented architectural patterns.
Audit Metadata