{"slug": "the-multi-agent-router-for-claude-code-drop-in-claude-setup-reel-route", "title": "The multi-agent router for Claude Code — drop-in .claude/ setup (reel: ROUTE)", "summary": "A developer released a free multi-agent routing setup for Claude Code that dispatches a single prompt to specialist subagents—research, build, write—running in parallel and merges the results. The drop-in configuration uses three files in the project's .claude/ folder and works with Claude Code's built-in subagent and Task tool without extra installation.", "body_md": "\n\n```\n---\nname: multi-agent-router\ndescription: A drop-in multi-agent routing setup for Claude Code — one prompt in, a router dispatches it to specialist subagents (research / build / write) running in parallel, then merges the result. Config + the exact .claude files. Triggers: \"multi-agent routing\", \"one prompt many agents\", \"claude subagent router\", \"route to specialists\".\n---\n```\n\nThis is the real routing setup I run in Claude Code— not pseudocode. You drop three files into your project's`.claude/`\n\nfolder and one prompt gets read, dispatched to the right specialists in parallel, and merged back into one result. Given away free by[@buildwith.conrad].\n\nMost people drive Claude one prompt at a time — ask, wait, ask again. But Claude Code can spawn\n**subagents** and run them in parallel. This setup adds a **router**: you give it one instruction,\nit decides which specialists the job needs (research the facts, build the code, write the copy),\nfires them at the same time, and stitches the outputs into one answer. You stop babysitting; the\nrouting is done for you.\n\n```\n        your one prompt\n              │\n          ┌───▼───┐\n          │ router │   reads intent → picks specialists\n          └─┬─┬─┬─┘\n        ┌───┘ │ └───┐\n   ┌────▼─┐ ┌─▼───┐ ┌▼─────┐\n   │research│ │build│ │write │   run in parallel\n   └────┬─┘ └─┬───┘ └┬─────┘\n        └───┐ │ ┌───┘\n          ┌─▼─▼─▼─┐\n          │ result │   merged into one answer\n          └───────┘\n```\n\nCreate these under your project root. That's the whole setup.\n\n```\n---\nname: researcher\ndescription: Gathers facts, reads the codebase/web, returns a tight findings brief. No code, no prose.\ntools: Read, Grep, Glob, WebSearch, WebFetch\nmodel: sonnet\n---\nYou are the research specialist. Given a task, collect only the facts needed to complete it —\nrelevant files, APIs, constraints, prior art. Return a compact bullet brief. Do not build or write copy.\n---\nname: builder\ndescription: Writes/edits the actual code for a scoped task. Returns a diff or the files changed.\ntools: Read, Edit, Write, Bash, Grep, Glob\nmodel: sonnet\n---\nYou are the build specialist. Given a scoped task + the researcher's brief, implement it cleanly,\nmatching existing conventions. Return what you changed and why. Do not do unrelated refactors.\n---\nname: writer\ndescription: Produces the human-facing copy — docs, UI text, release notes, captions.\ntools: Read, Grep, Glob\nmodel: sonnet\n---\nYou are the writing specialist. Given the task + context, produce clear, on-brand copy only.\nNo code. Match the voice of the surrounding material.\n---\ndescription: Read one request, dispatch to the right specialist subagents in parallel, merge the result.\n---\nYou are the ROUTER. Read the user's request below and:\n1. Decide which specialists it needs (researcher / builder / writer — one or several).\n2. Launch the chosen agents **in parallel** with the Task tool, each with a tight, scoped prompt\n   and the shared context they need.\n3. When they return, merge their outputs into ONE coherent result for the user — no raw dumps.\n\nOnly involve specialists the request actually needs. Request:\n\n$ARGUMENTS\n/route build a /pricing page: research our current plans, build the React page, write the headline + CTA\n```\n\nClaude reads the request, fires `researcher`\n\n+ `builder`\n\n+ `writer`\n\nat once, and hands you one\nfinished result instead of three round-trips. Add your own specialists (tester, reviewer, designer)\nby dropping more files in `.claude/agents/`\n\n.\n\nWorks in Claude Code out of the box — subagents + the Task tool are built in. No extra install.\n\nThis is one slice of how I run whole builds on autopilot — routing, parallel agents, one merged\noutput. If it was useful, the **Content-to-Cash** and **First Product** kits turn this kind of setup\ninto an actual sellable system: `buildwith.conrad`\n\nbio link.\n\n*Free giveaway from @buildwith.conrad — keyword ROUTE.*", "url": "https://wpnews.pro/news/the-multi-agent-router-for-claude-code-drop-in-claude-setup-reel-route", "canonical_source": "https://gist.github.com/conradcaffier03/10d8167f82d16a860ef6e939d9f943bd", "published_at": "2026-07-24 12:53:30+00:00", "updated_at": "2026-07-26 11:28:22.004804+00:00", "lang": "en", "topics": ["ai-agents", "developer-tools", "large-language-models"], "entities": ["Claude Code", "buildwith.conrad"], "alternates": {"html": "https://wpnews.pro/news/the-multi-agent-router-for-claude-code-drop-in-claude-setup-reel-route", "markdown": "https://wpnews.pro/news/the-multi-agent-router-for-claude-code-drop-in-claude-setup-reel-route.md", "text": "https://wpnews.pro/news/the-multi-agent-router-for-claude-code-drop-in-claude-setup-reel-route.txt", "jsonld": "https://wpnews.pro/news/the-multi-agent-router-for-claude-code-drop-in-claude-setup-reel-route.jsonld"}}