agent-teams-advanced
Pass
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of documentation and templates. No executable code or scripts are included in the skill package.
- [SAFE]: The skill instructs the agent on how to use platform-native multi-agent features. It outlines legitimate administrative patterns such as cost management, topology design (Hub-and-Spoke, Mesh), and git worktree isolation for parallel work.
- [NO_CODE]: The skill is purely informational and does not contain scripts, binaries, or executable logic, which eliminates typical execution-based attack vectors.
- [PROMPT_INJECTION]: The documented patterns describe workflows where agents process external, potentially untrusted data like pull request diffs. Ingestion points: PR diffs and source code files (referenced in examples/pr-review-board.md). Boundary markers: The documentation suggests using HTML comments (e.g., ) as delimiters. Capability inventory: The skill uses the Read, Glob, Grep, Bash, and Agent tools. Sanitization: No specific sanitization or filtering logic is implemented within the provided non-executable templates.
Audit Metadata