eval-agents/
Audit agent definitions for description specificity, model tier, tools scoping, and dispatch ambiguity between agents
When to use this
- When you have multiple agents in .claude/agents/ and want to detect dispatch ambiguity between them
- After adding new agents to an existing fleet to confirm descriptions are specific enough
- When an orchestrator consistently picks the wrong agent for a task
How to use
-
Open the template on GitHub and copy its content:
skills/eval-agents/SKILL.md -
Place the file inside
.claude/skills/in your project (or globally in~/.claude/to share across all projects). - Edit the Markdown content to match your project conventions, team constraints, and specific context.