# ModelPilot Execution Decision Engine

> ModelPilot Execution Decision Engine is a freemium MCP server for ai models and llm tooling. Policy-aware execution, verification and cost evidence for agent tasks. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Freemium
- **Platforms:** API, CLI
- **Website:** https://modelpilot.andycywu.cc/docs/agents
- **Repository:** https://github.com/andycywu/modelpilot
- **MCP Registry name:** io.github.andycywu/modelpilot
- **Categories:** AI models and LLM tooling
- **Listing updated:** 2026-11-15

## About

ModelPilot positions itself as a decision layer above your model providers rather than a model host or router. An agent hands it a task with privacy, risk, quality, latency and budget requirements, and ModelPilot picks a permitted execution path, runs it through your own targets, and checks the outcome.

The documented flow is to connect a provider credential and pass its live test, create an agent key with minimum execute scope and a small daily limit, then connect the remote MCP server. The page suggests telling the agent to use `modelpilot_execute_verified_task` and not to report completion unless the status is verified, returning the task ID, execution path, measured cost and verification result.

Provider keys stay inside ModelPilot. A native task API and an OpenAI-compatible API also exist, though the latter cannot express the full task and verification contract. A CLI is a developer preview installed from source.

## Features

- **Verified completion:** An agent is told to report done only when ModelPilot returns status verified.
- **Policy-aware routing:** Privacy, risk, quality, latency and budget requirements guide the chosen path.
- **Bring your own providers:** Execution runs through your credentials, which stay inside ModelPilot.
- **Cost and trajectory records:** Each task keeps decision, cost and verification evidence.
- **Scoped agent keys:** Keys carry minimum execute scope and a daily safety limit.

## Good for

- Make an agent prove a task finished before it says so
- Compare measured cost across permitted model paths
- Keep an audit trail of what ran for each task
- Cap an agent's daily spend while it works

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http modelpilot https://modelpilot.andycywu.cc/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://modelpilot.andycywu.cc/mcp
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "modelpilot": {
      "url": "https://modelpilot.andycywu.cc/mcp"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "modelpilot": {
      "type": "http",
      "url": "https://modelpilot.andycywu.cc/mcp"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "modelpilot": {
      "serverUrl": "https://modelpilot.andycywu.cc/mcp"
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "modelpilot": {
      "type": "http",
      "url": "https://modelpilot.andycywu.cc/mcp"
    }
  }
}
```

## Tools

- `modelpilot_execute_verified_task`: Run a task under policy and return a verified result with cost and path

## FAQ

### Is it a model host?

No. The site says it is not Ollama, a model host or another OpenRouter.

### Does the free tier cost anything?

The Free Sandbox gives 1,000 task decisions a month with your own provider keys and no SLA.

### How do I connect from VS Code?

Save a .vscode/mcp.json pointing at the remote URL with a bearer header that prompts for the mp_live key.

### What comes after the sandbox?

A paid outcome audit for one workflow, then a production control plane; prices are not listed.

### Is ModelPilot Execution Decision Engine free?

ModelPilot Execution Decision Engine has a free plan, with paid plans for more usage or features.

### How do I add ModelPilot Execution Decision Engine to Claude Code?

Run this command in your terminal: claude mcp add --transport http modelpilot https://modelpilot.andycywu.cc/mcp

### What are the best alternatives to ModelPilot Execution Decision Engine?

The closest alternatives to ModelPilot Execution Decision Engine are Claude Code Ultimate Guide, signafAIde, Toolkien, Lugh and HypePaper.

## Alternatives

- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [signafAIde](https://ai-powered.com/mcp/signafaide): signafAIde is a freemium MCP server for security.
- [Toolkien](https://ai-powered.com/mcp/toolkien): Toolkien is an open-source MCP server for ai models and llm tooling.
- [Lugh](https://ai-powered.com/mcp/lugh): Lugh is an open-source MCP server for ai models and llm tooling.
- [HypePaper](https://ai-powered.com/mcp/hypepaper): HypePaper is a freemium MCP server for ai models and llm tooling.
- [Agent Data Tools: Compact (find, describe, run)](https://ai-powered.com/mcp/agent-data-tools-compact-find-describe-run): Agent Data Tools: Compact (find, describe, run) is a freemium MCP server for developer tools.
- [Angel Agent Guard: pre-purchase check + prompt-injection scan](https://ai-powered.com/mcp/angel-agent-guard-pre-purchase-check-prompt-injection-scan): Angel Agent Guard: pre-purchase check + prompt-injection scan is a freemium MCP server for ai models and llm tooling.
- [Mizan Calibration](https://ai-powered.com/mcp/mizan-calibration): Mizan Calibration is a free MCP server for ai models and llm tooling.

---
Source: https://ai-powered.com/mcp/modelpilot-execution-decision-engine
