mcp-agno
# mcp-agno
An MCP (Model Context Protocol) server that gives AI assistants direct access to the [AGNO](https://docs.agno.com) framework documentation.
Instead of crawling docs manually or relying on outdated training data, Claude and other MCP-compatible clients can browse, fetch, and search AGNO docs in real time.
## What it does
Exposes three MCP tools:
| Tool | Description |
|------|-------------|
| `list_agno_sections` | List all available documentation sections with their slugs and URLs |
| `get_agno_page` | Fetch the full content of a documentation page by slug or URL |
| `search_agno_docs` | Keyword search across the catalog — returns matching slugs without fetching pages |
On startup the server seeds a catalog of core sections (Introduction, Agents, Teams, Workflows, AgentOS) and auto-discovers additional pages from the live AGNO sitemap.
---
## Requirements
- Python 3.11+
- [uv](https://docs.astral.sh/uv/) (recommended) or pip
---
## Installation
### With uv (recommended)
```bash
git clone https://github.com/Attilio81/MCP_AGNO.git
cd MCP_AGNO
uv pip install -e .
```
### With pip
```bash
git clone https://github.com/Attilio81/MCP_AGNO.git
cd MCP_AGNO
pip install -e .
```
This installs the `mcp-agno` command.
---
## Usage with Claude Desktop
Add the server to your Claude Desktop config (`claude_desktop_config.json`):
```json
{
"mcpServers": {
"agno-docs": {
"command": "mcp-agno"
}
}
}
```
If you installed with uv in a virtual environment, use the full path:
```json
{
"mcpServers": {
"agno-docs": {
"command": "/path/to/venv/bin/mcp-agno"
}
}
}
```
Restart Claude Desktop. The three tools will appear automatically.
---
## Usage with Claude Code (CLI)
```bash
claude mcp add agno-docs mcp-agno
```
Or add it to your project's `.mcp.json`:
```json
{
"mcpServers": {
"agno-docs": {
"command": "mcp-agno"
}
}
}
```
---
## Running manually (for testing)
```bash
mcp-agno
```
The server runs over stdio and waits for MCP messages.
---
## Development
```bash
git clone https://github.com/Attilio81/MCP_AGNO.git
cd MCP_AGNO
uv pip install -e ".[dev]"
pytest
```
---
## Resources
- [AGNO Documentation](https://docs.agno.com) — official docs browsed by this server
- [AGNO Cookbook](https://github.com/agno-agi/agno/tree/main/cookbook) — practical examples and ready-to-run recipes from the AGNO team
---
## License
MIT
TDQS
Scored across 3 tools
Each tool has a clearly distinct purpose: get_agno_page retrieves specific page content, list_agno_sections enumerates available sections, and search_agno_docs performs keyword searches. There is no overlap in functionality, making tool selection straightforward for an agent.
All tool names follow a consistent verb_noun pattern with 'agno' as a prefix (get_agno_page, list_agno_sections, search_agno_docs). This uniformity enhances readability and predictability across the toolset.
With 3 tools, the count is slightly low but reasonable for a documentation-focused server. It covers core operations (fetch, list, search), though additional tools like update or delete might be expected in a broader CRUD context, but are unnecessary here given the domain.
The toolset provides complete coverage for interacting with AGNO documentation: listing sections, fetching specific pages, and searching content. There are no obvious gaps for the stated purpose, and agents can perform typical documentation workflows without dead ends.