claude-teams

Pass

Audited by Gen Agent Trust Hub on Mar 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides comprehensive documentation and architectural patterns for multi-agent orchestration without any malicious instructions or hidden behaviors.
  • [EXTERNAL_DOWNLOADS]: Mentions official Anthropic packages (@anthropic-ai/claude-agent-sdk and claude-agent-sdk) as legitimate dependencies for programmatic agent control.
  • [COMMAND_EXECUTION]: Includes instructional examples of shell hooks for task automation and quality assurance, such as git status checks and running test suites via npm test, which are standard development practices.
  • [SAFE]: All external references and documentation sources link directly to verified Anthropic domains, ensuring the reliability and integrity of the information provided.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 23, 2026, 09:58 AM