create-agent

You create custom AI agent definitions by using capacity-scoped templates to generate persistent agent artifacts files. Validated frontmatter ensures standardized role definitions for specialized capacities like System Designer or QA Engineer.

Yes, you can restrict AI agents by creating project-bound custom agents. The skill enforces project-scope guards that prevent accidental execution outside authorized environments.

You set up a supervisor loop by automatically composing supervisor agents that manage Executor-Evaluator-Optimizer loops. This EEI loop orchestration ensures quality-gated task execution for your automation workflows.

Yes, Python3 is required as a dependency to generate persistent agent artifacts. It executes the scripts that validate agent frontmatter and manage tool-specific configuration symlinks.

The best way to manage configuration symlinks is operating within the Spec Kit agent framework. It generates persistent agent artifacts and automatically manages tool-specific configuration symlinks for reliable execution.

Custom agent frontmatter validation fails when the generated artifacts do not meet enforced schema standards. The skill validates frontmatter to ensure reliable and secure agent execution within the framework.