# apiwatch MCP server

> Monitors API, model, and MCP server changes to identify breaking updates and suggest code fixes via GitHub Actions or CLI.

- Repository: https://github.com/ReignVentures/apiwatch
- Page: https://hostedmcp.org/servers/reignventures/apiwatch
- Transport: remote
- Stars: 0
- License: MIT
- Latest version: 1.1.0

## Install

### Claude Code

```sh
claude mcp add --transport http apiwatch https://apiwatch.reignventures.co/mcp
```

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

```json
{
  "mcpServers": {
    "apiwatch": {
      "url": "https://apiwatch.reignventures.co/mcp"
    }
  }
}
```

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

```json
{
  "servers": {
    "apiwatch": {
      "type": "http",
      "url": "https://apiwatch.reignventures.co/mcp"
    }
  }
}
```

### Codex (`~/.codex/config.toml`)

```toml
[mcp_servers.apiwatch]
url = "https://apiwatch.reignventures.co/mcp"
```

### Gemini CLI (`~/.gemini/settings.json`)

```json
{
  "mcpServers": {
    "apiwatch": {
      "httpUrl": "https://apiwatch.reignventures.co/mcp"
    }
  }
}
```

## FAQ

### How does apiwatch identify breaking changes in my code?

It scans your repository for specific call sites, using token-aware matching to distinguish between similar model IDs and following constants and notebooks.

### Does my code leave my machine during a local scan?

No, scans performed via the CLI run locally with no network calls, ensuring your code remains private.

### How are the API change records verified?

Records are verified by a reviewer who did not write them, checking against the vendor's primary source, and confirmed by an automated check.

### Can I use apiwatch with AI agents?

Yes, you can add the apiwatch MCP server to your client to allow agents to query for model and API changes during their workflow.