model-routing-preset-builder
Pass
Audited by Gen Agent Trust Hub on Aug 16, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses a sibling Python script located at
<model-capability-router>/scripts/route.pyto validate generated TOML presets. This is a standard local tool execution pattern for development-oriented skills. - [SAFE]: The instructions include explicit security requirements for 'Gates.' Capabilities that involve mutating the host, accessing secrets, deploying, publishing, or financial transactions require an explicit gate independent of model selection.
- [SAFE]: The skill implements a human-in-the-loop requirement, explicitly instructing the agent to ask the operator to approve the preset before it becomes a default.
- [SAFE]: Best practices for security and stability are followed, such as recording capability requirements by stable names rather than local filesystem paths to prevent path leakage or traversal risks.
Audit Metadata