agent-creator
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides structured instructions and templates for generating agent definitions. It outlines clear development phases, including discovery, requirement resolution, writing, placement, and validation.
- [EXTERNAL_DOWNLOADS]: The workflow includes a validation step using
uvx skilllint@latest. This is a standard procedure for linting agent files to ensure they conform to the required YAML schema and platform constraints. - [COMMAND_EXECUTION]: The skill suggests running validation commands (
skilllint,claude plugin validate) and local project scripts (uv run check_agent_auto_discovery.py). These commands are typical for software development lifecycles and are used here to maintain artifact quality.
Audit Metadata