agent-finder

Inventory and match subagents across project, user, and built-in scopes.

Updated May 12, 2026
One-click install
npx skills add https://github.com/RorySullivan1/claudeBrain --skill agent-finder-rorysullivan1
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-finder
Source: https://github.com/RorySullivan1/claudeBrain/tree/main/example-project/.claude/skills/agent-finder
Command: npx skills add https://github.com/RorySullivan1/claudeBrain --skill agent-finder-rorysullivan1

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Find and delegate to the optimal Claude Code subagent(s) for a task instead of defaulting to general-purpose or doing everything in the main thread. Use whenever a task could be delegated — research, codebase exploration, code review, test runs, domain-specific work, or anything that would flood the main context with output — and whenever the user asks which agent to use, to "use the right agent", to run work in parallel, or to spin up a specialist. Trigger even if the user doesn't say the word "agent": any multi-step or high-volume side task is a delegation decision.

Core Features & Use Cases

  • Inventory and match real-time subagents from project and user scopes, plus three built-ins (Explore, Plan, general-purpose), to surface the best-fit agent for the task.
  • Follow a four-step delegation workflow (inventory → match → choose topology → delegate) to decide single vs parallel execution and to formulate a crisp delegation prompt.
  • Provide concrete guidance on when to use each agent, how to inspect their capabilities, and how to tie results back to the main task.

Quick Start

Ask the agent-finder to identify the best-fit subagent for the current task and present a concise delegation plan.

Frequently Asked Questions about agent-finder

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I delegate tasks to subagents in Claude Code?

Delegate tasks by inventorying available project, user, and built-in subagents, matching the best fit to your task, choosing single or parallel execution topology, and generating a delegation prompt with a return contract.

When should I use subagents instead of the main thread for code work?

Use subagents for multi-step or high-volume side tasks like research, codebase exploration, code review, and test runs that would flood the main context with output, keeping your primary thread clean and focused.

What's the best way to run parallel subagents for codebase exploration?

The best way is applying the delegation workflow: inventory available subagents, match them to task segments, choose a parallel execution topology, and craft crisp delegation prompts with constraints for each agent.

Can I inspect subagent capabilities before delegating a task?

Yes, you can inspect subagent capabilities during the inventory and match phases, which surface real-time subagents from project and user scopes plus built-ins, providing concrete guidance on when to use each agent.

How do I formulate a return contract when delegating to a subagent?

Formulate a return contract by defining concise constraints and expected output formats within the delegation prompt, ensuring the chosen subagent ties its results directly back to your main task objective.

Does this work with built-in general-purpose agents or only custom subagents?

It works with both, inventorying and matching across project scopes, user scopes, and three built-in agents (Explore, Plan, general-purpose) to surface the best-fit agent for any given task.