agentleFS
armenrarmenr/modelmux/.claude/skills/getting-started/SKILL.md

getting-started

Skill2 starsChanged 4 months agoMITView on GitHub

Use when someone has just cloned modelmux and wants to plug in OpenRouter and see heterogeneous routing working. Walks clone → install → key → run → verify, checking each step.

  • Pipes a download into a shell
  • Reads credentials

32 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 32 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If it's missing, either curl -fsSL https://bun.sh/install | bash or use DevBox (devbox shell).

    line 23

“Common” is measured: kept by at least 6 in 10 of the 32 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Getting started with modelmux

Walk the user through these steps in order. Run the checks yourself where you can, and confirm each step before moving on. Never print the value of an API key.

This is the from-source path (clone the repo, run with Bun). Running a prebuilt binary instead? Skip install and run ./modelmux — see the README. The fuller dev workflow is in docs/development.md.

1. Prerequisites

Confirm Bun is installed:

bun --version

If it's missing, either curl -fsSL https://bun.sh/install | bash or use DevBox (devbox shell). Then install dev dependencies:

bun install

2. Plug in your OpenRouter key

The proxy needs an OpenRouter key for any non-Claude route (https://openrouter.ai/keys).

cp .env.example .env
# then edit .env and set OPENROUTER_API_KEY=sk-or-v1-...

.env is gitignored — never commit it.

3. Start the proxy

bun run proxy

You should see modelmux listening on http://localhost:8787. Leave it running in its own terminal.

4. Opt Claude Code into the proxy

Routing only happens when Claude Code points at the proxy:

cp .claude/settings.json.example .claude/settings.json

This sets ANTHROPIC_BASE_URL=http://127.0.0.1:8787. Restart Claude Code — that env var is read once at startup, so an already-running session won't pick it up.

5. See it route

In a fresh Claude Code session (proxy running, settings enabled), dispatch a subagent — e.g. the bundled glm-researcher (tagged <<route:flagship>>) — then inspect the decision log:

tail -n 5 decisions.jsonl

You want to see the subagent land on OpenRouter and the main loop stay on Claude:

{ "isSubagent": true,  "matchedRule": "tag:flagship", "upstream": "openrouter", "resolvedModel": "z-ai/glm-5.2" }
{ "isSubagent": false, "matchedRule": "default",      "upstream": "anthropic",  "resolvedModel": "passthrough" }

6. (Optional) confirm the model slugs are live

bin/mux check-latest

Shortcut

You can also dispatch the setup-assistant agent to auto-check Bun, the key, config validity, and the proxy, and report what's left.

Troubleshooting

  • OPENROUTER_API_KEY is not set (HTTP 400) → set the key in .env (step 2).
  • Connection refused → the proxy isn't running (step 3).
  • Subagent still on Claude → you didn't restart Claude Code after step 4.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.