How do I install the Openclaw Model Router skill?

Run npx skillmds@latest add mnfst/openclaw-model-router in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Openclaw Model Router safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Openclaw Model Router?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Openclaw Model Router free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Openclaw Model Router?

mnfst (@mnfst) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Openclaw Model Router by mnfst — Agent Skill

Name: Openclaw Model Router
Author: mnfst

Openclaw Model Router

Install and configure the Manifest LLM router plugin for OpenClaw. Use when the user says "/install-manifest", "install manifest", "add manifest plugin", "setup model router", "reduce LLM costs", or wants to install the Manifest plugin to route requests, cut costs, add fallbacks, and set spending limits. Accepts an optional mode (local or cloud) and API key.

mnfst aeff678 2 files · 6.6 KB Updated Sep 17, 2026

File contents

OpenClaw Model Router

Install the Manifest plugin for OpenClaw — an open-source LLM router that saves up to 70% on model costs.

Why Manifest

Save costs — Routes each request to the cheapest model that can handle it. A 23-dimension scoring algorithm picks the right tier in under 2ms.
Route LLM requests — Supports 300+ models across Anthropic, OpenAI, Google, DeepSeek, xAI, Mistral, and more. One endpoint, all providers.
Add fallbacks — If a model fails or hits a rate limit, traffic shifts to an alternative automatically. No downtime, no code changes.
Set limits — Configure spending caps and email alerts per hour, day, week, or month.

Workflow

1. Determine parameters

Mode (optional, default local):
- local — All data stays on your machine. No account needed. Dashboard at http://127.0.0.1:2099.
- cloud — Dashboard at app.manifest.build. Requires an API key.
Key (required for cloud mode): A mnfst_* API key from app.manifest.build.

If the user doesn't specify a mode, default to local. If cloud mode and no key, ask for it.

2. Run the install script

# Local mode (default — no account needed):
bash skills/openclaw-model-router/scripts/install_manifest.sh

# Cloud mode:
bash skills/openclaw-model-router/scripts/install_manifest.sh --mode cloud --key mnfst_YOUR_KEY

Use --dry-run to preview changes without modifying anything.

3. Show status table

After installation, run the diagnostic table:

bash skills/manifest-status/scripts/manifest_status.sh

Output the table exactly as printed. No extra commentary.

4. Next steps

Tell the user:

Dashboard is available at http://127.0.0.1:2099 (local) or app.manifest.build (cloud)
The default model is now manifest/auto — requests are routed based on complexity tier
Customize routing tiers and spending limits in the dashboard

---name: openclaw-model-routerdescription: Install and configure the Manifest LLM router plugin for OpenClaw. Use when the user says "/install-manifest", "install manifest", "add manifest plugin", "setup model router", "reduce LLM costs", or wants to install the Manifest plugin to route requests, cut costs, add fallbacks, and set spending limits. Accepts an optional mode (local or cloud) and API key.---# OpenClaw Model RouterInstall the [Manifest](https://github.com/mnfst/manifest) plugin for OpenClaw — an open-source LLM router that saves up to 70% on model costs.## Why Manifest- **Save costs** — Routes each request to the cheapest model that can handle it. A 23-dimension scoring algorithm picks the right tier in under 2ms.- **Route LLM requests** — Supports 300+ models across Anthropic, OpenAI, Google, DeepSeek, xAI, Mistral, and more. One endpoint, all providers.- **Add fallbacks** — If a model fails or hits a rate limit, traffic shifts to an alternative automatically. No downtime, no code changes.- **Set limits** — Configure spending caps and email alerts per hour, day, week, or month.## Workflow### 1. Determine parameters- **Mode** (optional, default `local`): - `local` — All data stays on your machine. No account needed. Dashboard at `http://127.0.0.1:2099`. - `cloud` — Dashboard at [app.manifest.build](https://app.manifest.build). Requires an API key.- **Key** (required for cloud mode): A `mnfst_*` API key from [app.manifest.build](https://app.manifest.build).If the user doesn't specify a mode, default to `local`. If cloud mode and no key, ask for it.### 2. Run the install script```bash# Local mode (default — no account needed):bash skills/openclaw-model-router/scripts/install_manifest.sh# Cloud mode:bash skills/openclaw-model-router/scripts/install_manifest.sh --mode cloud --key mnfst_YOUR_KEY```Use `--dry-run` to preview changes without modifying anything.### 3. Show status tableAfter installation, run the diagnostic table:```bashbash skills/manifest-status/scripts/manifest_status.sh```Output the table exactly as printed. No extra commentary.### 4. Next stepsTell the user:- Dashboard is available at **http://127.0.0.1:2099** (local) or **app.manifest.build** (cloud)- The default model is now `manifest/auto` — requests are routed based on complexity tier- Customize routing tiers and spending limits in the dashboard

Frequently asked questions