Metadata-Version: 2.4
Name: agentvee-mcp
Version: 0.1.3
Summary: MCP server for AgentVee — file uploads for AI agents
Project-URL: Homepage, https://agentvee.io
Project-URL: Repository, https://github.com/agentvee/agentvee
License-Expression: MIT
Keywords: agentvee,ai-agents,file-upload,mcp
Classifier: Development Status :: 3 - Alpha
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Requires-Python: >=3.11
Requires-Dist: httpx>=0.27.0
Requires-Dist: mcp>=1.0.0
Description-Content-Type: text/markdown

# agentvee-mcp

MCP server for [AgentVee](https://agentvee.io) — file uploads for AI agents with local `filePath` support.

> **Note:** This is currently a test version. Full launch coming soon.

## Quick Start

1. **Get an API key** at [agentvee.io/dashboard](https://agentvee.io/dashboard)

1. **Add to your MCP config** (e.g. `~/.cursor/mcp.json`):

```json
{
  "mcpServers": {
    "agentvee": {
      "command": "uvx",
      "args": ["agentvee-mcp"],
      "env": {
        "AGENTVEE_API_KEY": "your-api-key-here"
      }
    }
  }
}
```

Or with `pip`:

```json
{
  "mcpServers": {
    "agentvee": {
      "command": "python",
      "args": ["-m", "agentvee_mcp"],
      "env": {
        "AGENTVEE_API_KEY": "your-api-key-here"
      }
    }
  }
}
```

1. **Restart your MCP client** (Cursor, Claude Desktop, etc.)

## Installation

```bash
pip install agentvee-mcp
```

Or with uv:

```bash
uvx agentvee-mcp --help
```

## Supported Clients

Any MCP client that supports stdio transport:

- [Cursor](https://cursor.sh)
- [Claude Desktop](https://claude.ai/download)
- [Cline](https://github.com/cline/cline)
- [Continue](https://continue.dev)

## Available Tools

| Tool                | Description                                          |
| ------------------- | ---------------------------------------------------- |
| `upload_and_wait`   | Upload a file and wait for a shareable download link |
| `upload_file`       | Upload a file (returns upload ID for status polling) |
| `upload_from_url`   | Upload a file from a public URL                      |
| `get_upload_status` | Check the status of an upload                        |
| `get_download_url`  | Get the download URL for a completed upload          |

### Upload Methods

Each upload tool supports three mutually exclusive input methods:

- **`filePath`** — Local file path (stdio only, most efficient for local files — zero tokens)
- **`url`** — Public URL for the server to fetch
- **`content`** — Base64-encoded file content (fallback, uses tokens)

## Environment Variables

| Variable                 | Required | Default                                 | Description            |
| ------------------------ | -------- | --------------------------------------- | ---------------------- |
| `AGENTVEE_API_KEY`       | Yes      | —                                       | Your AgentVee API key  |
| `AGENTVEE_API_BASE_URL`  | No       | `https://agentvee-api-develop.fly.dev`  | API base URL           |

## CLI

```bash
agentvee-mcp --help
agentvee-mcp --version
```

## HTTP Transport

To run as an HTTP server (Streamable HTTP):

```bash
agentvee-mcp-http
```

Or:

```bash
python -m agentvee_mcp.http
```

## Also Available

- **Node.js:** `npx -y @agentvee/mcp` ([npm](https://www.npmjs.com/package/@agentvee/mcp))

## License

MIT
