# FetchSandbox

> FetchSandbox is an MCP server for developer tools. Rehearse AI workflows against stateful API twins and get a pass or fail receipt. It can run locally on your machine or as a hosted remote server.

- **Kind:** MCP server
- **Platforms:** API, CLI, Web
- **Website:** https://fetchsandbox.com
- **Repository:** https://github.com/fetchsandbox/mcp
- **MCP Registry name:** io.github.fetchsandbox/mcp
- **Categories:** [developer tools](https://ai-powered.com/categories/developer-tools)
- **Listing updated:** 2026-11-18

## About

FetchSandbox is built for the moment before an agent or n8n workflow touches production. It replays your workflow against isolated twins of third-party APIs and shows the stored side effects, so you see what the workflow actually did when retries, lost responses, concurrent edits or partial failures happen.

The catalogue includes Stripe, Slack, Gmail, HubSpot, Twilio, Shopify Admin, GitHub, PayPal, Plaid and many more. You agree the rule first, for example whether a retry after a lost response may duplicate an action, and the receipt compares expected with actual behaviour.

The hosted endpoint covers planning and verification, while find_bugs, fix_bug and prove_fix read your project from disk and so run only in the local npx fetchsandbox-mcp server. The registry describes proving a fix: failing on the old code and passing on the new. An API key is needed for planning, verification and managed tunnels; discovery is anonymous.

## Features

- **API twins:** Stateful stand-ins for dozens of provider APIs.
- **Failure scenarios:** Test retries, lost responses and state changes mid-run.
- **Receipts:** Expected versus actual, marked passed, failed or could not verify.
- **Fix proving:** Show a fix fails on old code and passes on new.
- **Local code tools:** find_bugs, fix_bug and prove_fix work on your project.
- **No live keys:** Twin rehearsals run without provider credentials.

## Good for

- Test webhook retry behaviour before launch
- Verify an n8n workflow handles duplicates
- Prove a bug fix with before and after runs
- Check integration code against a Stripe twin

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http mcp https://fetchsandbox.com/mcp/v1
```

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

```sh
https://fetchsandbox.com/mcp/v1
```

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

```json
{
  "mcpServers": {
    "mcp": {
      "url": "https://fetchsandbox.com/mcp/v1"
    }
  }
}
```

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

```json
{
  "servers": {
    "mcp": {
      "type": "http",
      "url": "https://fetchsandbox.com/mcp/v1"
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "mcp": {
      "serverUrl": "https://fetchsandbox.com/mcp/v1"
    }
  }
}
```

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

```json
{
  "mcpServers": {
    "mcp": {
      "type": "http",
      "url": "https://fetchsandbox.com/mcp/v1"
    }
  }
}
```

## Tools

- `validate_integration`: Starting point for checking an integration.
- `list_workflows`: Lists available workflows.
- `run_workflow`: Runs a workflow against API twins.
- `verify_behavior`: Checks workflow behaviour against agreed rules.
- `find_bugs`: Analyses a local project for bugs (local server).
- `fix_bug`: Applies a fix to a found bug (local server).
- `prove_fix`: Shows a fix fails before and passes after (local server).

## FAQ

### Do I need provider keys?

The site says twin rehearsals run without live provider keys.

### Why do some tools need local install?

They read your project from disk, which a hosted server cannot do.

### Is there an API key?

Yes, FETCHSANDBOX_API_KEY is needed for planning and verification.

### Can I turn off telemetry?

Set FETCHSANDBOX_TELEMETRY to 0.

### How do I add FetchSandbox to Claude Code?

Run this command in your terminal: claude mcp add --transport http mcp https://fetchsandbox.com/mcp/v1

### What are the best alternatives to FetchSandbox?

The closest alternatives to FetchSandbox are REA, Context7, Chrome DevTools MCP, Codebase Memory and GitHub.

## Where this comes from

- Description: written from the project's README and the project's website.
- Install steps: from the official MCP Registry entry.

## Alternatives

- [REA](https://ai-powered.com/mcp/rea): REA is an open-source MCP server for developer tools.
- [Context7](https://ai-powered.com/mcp/context7): Context7 is a freemium MCP server for developer tools.
- [Chrome DevTools MCP](https://ai-powered.com/mcp/chrome-devtools-mcp-2): Chrome DevTools MCP is an open-source MCP server for browser automation.
- [Codebase Memory](https://ai-powered.com/mcp/codebase-memory): Codebase Memory is an open-source MCP server for knowledge and memory.
- [GitHub](https://ai-powered.com/mcp/github-2): GitHub is an open-source MCP server for developer tools.
- [Serena MCP: the IDE for your agent](https://ai-powered.com/mcp/serena-mcp-the-ide-for-your-agent): Serena MCP: the IDE for your agent is an open-source MCP server for developer tools.
- [awesome-mcp-servers](https://ai-powered.com/mcp/awesome-mcp-servers-2): awesome-mcp-servers is an open-source MCP server for developer tools.
- [XERJ](https://ai-powered.com/mcp/xerj): XERJ is an open-source MCP server for knowledge and memory.

---
Source: https://ai-powered.com/mcp/fetchsandbox
