# docvet MCP server

> Analyzes Python docstrings for presence, completeness, freshness, and rendering compatibility to improve AI agent context.

- Repository: https://github.com/Alberto-Codes/docvet
- Page: https://hostedmcp.org/servers/alberto-codes/docvet
- Transport: stdio
- Stars: 10
- License: MIT
- Latest version: 3.0.1

## Install

### Claude Code

```sh
claude mcp add docvet -- uvx docvet
```

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

```json
{
  "mcpServers": {
    "docvet": {
      "command": "uvx",
      "args": [
        "docvet"
      ]
    }
  }
}
```

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

```json
{
  "servers": {
    "docvet": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "docvet"
      ]
    }
  }
}
```

### VS Code CLI

```sh
code --add-mcp '{"name":"docvet","type":"stdio","command":"uvx","args":["docvet"]}'
```

### Claude Desktop (`claude_desktop_config.json`)

```json
{
  "mcpServers": {
    "docvet": {
      "command": "uvx",
      "args": [
        "docvet"
      ]
    }
  }
}
```

### Codex

```sh
codex mcp add docvet -- uvx docvet
```

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

```toml
[mcp_servers.docvet]
command = "uvx"
args = ["docvet"]
```

### Gemini CLI

```sh
gemini mcp add -s user docvet uvx -- docvet
```

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

```json
{
  "mcpServers": {
    "docvet": {
      "command": "uvx",
      "args": [
        "docvet"
      ]
    }
  }
}
```

## FAQ

### What does docvet check for?

It performs checks for docstring presence, enrichment (completeness), freshness (staleness signals via git), rendering compatibility with Griffe, and package visibility.

### Can docvet prove that my docstrings are semantically correct?

No, it provides structural and freshness heuristics but does not provide semantic proof that the prose describes runtime behavior correctly.

### How can I use docvet with an AI agent?

You can start an MCP server specifically for AI agent integration using the command `docvet mcp`.
