agents-swarm-orchestration
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a comprehensive guide for multi-agent orchestration and contains no executable malicious code, obfuscation, or persistence mechanisms.
- [PROMPT_INJECTION]: No malicious injection patterns were found. The skill includes a dedicated section on 'Security: Dynamic Context in Prompts' that provides specific advice on validating file paths and sanitizing untrusted inputs to prevent agent hijacking.
- [DATA_EXFILTRATION]: No patterns for accessing sensitive files or exfiltrating data were detected. References to external URLs in SKILL.md and data/sources.json point to reputable documentation, research papers (ArXiv), and official service domains (OpenAI, Anthropic).
- [REMOTE_CODE_EXECUTION]: The skill describes the use of standard subagent tools and isolation methods like Git worktrees and sandboxed containers but does not contain commands for downloading or executing untrusted remote scripts.
Audit Metadata