> ## Documentation Index
> Fetch the complete documentation index at: https://docs.doconda.com/llms.txt
> Use this file to discover all available pages before exploring further.

# CrewAI

> Doconda como herramientas de tus agentes de CrewAI.

```bash theme={null}
uv add crewai mcp
```

```python theme={null}
import os
from crewai import Agent, Crew, Task
from crewai.mcp import MCPServerHTTP

key = os.environ["DOCONDA_API_KEY"]

writer = Agent(
    role="Redactor",
    goal="Entregar documentos impecables",
    backstory="Escribes informes y contratos claros.",
    mcps=[MCPServerHTTP(url="https://api.eu.doconda.com/mcp", headers={"Authorization": f"Bearer {key}"}, streamable=True)],
)

task = Task(
    description="Prepara un informe en Word sobre las ventas del trimestre y devuelve el enlace.",
    expected_output="El enlace de descarga del Word.",
    agent=writer,
)

Crew(agents=[writer], tasks=[task]).kickoff()
```

Para usarlo en local, en vez del remoto:

```python theme={null}
from crewai.mcp import MCPServerStdio

MCPServerStdio(command="npx", args=["-y", "@doconda/mcp"], env={"DOCONDA_API_KEY": key})
```


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.