swarm-operations
Pass
Audited by Gen Agent Trust Hub on Mar 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is strictly documentation for swarm operations, providing guidance on using TeamCreate, SendMessage, and Agent tools.
- [DATA_EXPOSURE]: The file references local paths such as
~/.claude/teams/and~/.claude/tasks/. These are standard internal directories used by the platform for storing team configuration and task state, and their mention is for debugging and transparency purposes. - [COMMAND_EXECUTION]: The documentation includes an example of a
permission_requestmessage containing aBashtool call (npm install). This is provided as a schema example for inter-agent communication and does not execute any commands on the host system. - [PROMPT_INJECTION]: No instructions were found that attempt to override system prompts, bypass safety filters, or extract sensitive instructions.
Audit Metadata