team-setup
Pass
Audited by Gen Agent Trust Hub on Feb 21, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- Indirect Prompt Injection (LOW): The skill processes untrusted input (user requirements, project files) to formulate prompts for sub-agents. * Ingestion points: User requirements, CLAUDE.md, and directory structures. * Boundary markers: Absent in the provided prompt templates. * Capability inventory: Spawns new agents via TeamCreate/Task and writes configuration files. * Sanitization: No sanitization or validation of the input data before interpolation.
- Dynamic Execution (LOW): The skill instructs the agent to create and modify .claude/agents/{role}.md files, which define the logic and permissions of sub-agents based on dynamic roles.
Audit Metadata