subagent-teams
Pass
Audited by Gen Agent Trust Hub on Mar 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security threats were detected. The skill follows best practices for task orchestration and context management within the AI agent environment.
- [PROMPT_INJECTION]: Analysis identified a standard surface for indirect prompt injection common to orchestration skills. 1. Ingestion points: Data from the local codebase and user-provided requirements are ingested into the subagent context during the exploration and implementation phases defined in SKILL.md and references/prompt-templates.md. 2. Boundary markers: The skill uses structured templates in references/prompt-templates.md to distinguish between instructions and context data. 3. Capability inventory: Subagents are permitted to use system tools such as Bash, Read, Grep, and Glob for file manipulation and testing, as outlined in the Team Assignment phase of SKILL.md. 4. Sanitization: No specialized sanitization or validation of content retrieved from the codebase is explicitly defined in the provided files.
Audit Metadata