How to Use the Haystack (deepset Cloud) MCP in AutoGen

Q: Can one AutoGen agent verify the work of another using this MCP server?

Absolutely. One agent can perform an action with runpipeline, and a second agent can use getpipeline to check its configuration or searchdocuments to find alternative evidence. This adversarial setup is a core strength of AutoGen.

Q: How do I assign specific Haystack (deepset Cloud) tools to different agents in AutoGen?

When you initialize your AssistantAgent, you pass a list of functions to its llmconfig. You can create separate lists of tools from this MCP server and give one list to your 'executor' agent and a different list to your 'critic' agent.

Q: What's a good use case for listpipelines in an AutoGen setup?

An 'Architect' agent could use listpipelines to review all available pipelines in a workspace. It could then recommend the best one for a specific task to an 'Executor' agent, kicking off the workflow.

Equip your AutoGen agent teams with Haystack tools. Let them debate, verify, and execute RAG pipelines to reach a consensus.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Haystack (deepset Cloud) MCP to AutoGen

Create your Vinkius account to connect Haystack (deepset Cloud) to AutoGen and route execution through our secure gateway. The platform manages server hosting, runtime updates, and security layers. Configuration requires no manual server provisioning.

GDPR Free for Subscribers

Setup Haystack (deepset Cloud) with AutoGen

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Enable Agent-Based Pipeline Validation

Create a team of agents that check each other's work. One agent, the 'Executor', can use `run_pipeline` to get results for a query. A second agent, the 'Auditor', can then use `get_pipeline` to inspect the exact configuration that was used. The Auditor can flag potential issues, like an outdated model or a poorly configured node. This conversational approach, where agents debate the validity of a pipeline run, leads to more reliable and well-reasoned outcomes.

Assign Specialized Roles to Your AutoGen Agents

Build a multi-agent system where each agent has a specific job related to Haystack. A 'WorkspaceManager' agent could be given only the `list_workspaces` tool to decide which environment is appropriate for a task. A 'SearchSpecialist' agent would have access to `search_documents` to find information within that workspace. This separation of concerns mirrors how human teams operate. It prevents a single agent from having too much authority and allows for more complex, orchestrated workflows. This MCP server provides the granular tools needed for this kind of design.

Ground Debates in Factual Source Material

Stop agents from hallucinating during a debate. You can create a 'Librarian' agent whose only job is to provide facts using the `list_files` and `get_file` tools. When other agents are discussing a topic, they can ask the Librarian to pull up the original source document from your Haystack index. This forces the conversation to stay grounded in the data you've provided. The agents' conclusions are based on shared, verifiable information, making their final output much more trustworthy.

Setup guide

Set up Haystack (deepset Cloud) MCP in AutoGen

Prerequisites

Python 3.10+ installed
autogen-ext[mcp] package
Active Vinkius subscription with a valid endpoint token

1

Install AutoGen with MCP
Run pip install "autogen-ext[mcp]" autogen-agentchat. The MCP extension includes mcp_server_tools for stateless tool access.
2

Fetch tools from the MCP
Call mcp_server_tools(SseServerParams(url=...)) with your Vinkius endpoint. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com.
3

Run your agent
Pass the tools to AssistantAgent and call agent.run(). The agent invokes Haystack (deepset Cloud) tools and returns structured results.

agent.py

from autogen_ext.tools.mcp import SseServerParams, mcp_server_tools
from autogen_agentchat.agents import AssistantAgent
from autogen_ext.models.openai import OpenAIChatCompletionClient

server_params = SseServerParams(
    url="https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
)

tools = await mcp_server_tools(server_params)

agent = AssistantAgent(
    name="Haystack (deepset Cloud)_assistant",
    model_client=OpenAIChatCompletionClient(model="gpt-4o"),
    tools=tools,
)

result = await agent.run("List recent Haystack (deepset Cloud) data")
print(result.messages[-1].content)

Get your connection token →

Prerequisites

Python 3.10+ installed
autogen-ext[mcp] + autogen-agentchat
Active Vinkius subscription with a valid endpoint token

1

Install dependencies
Same packages as above. McpWorkbench is ideal when your agent needs stateful sessions across multiple tool calls.
2

Use McpWorkbench as context manager
Wrap your agent in async with McpWorkbench(...) to maintain shared state and resources. The workbench manages the full MCP session lifecycle.
3

Run with workbench
Pass workbench=workbench to your agent. State is preserved across multiple tool calls within the same session.

agent.py

from autogen_ext.tools.mcp import McpWorkbench, SseServerParams
from autogen_agentchat.agents import AssistantAgent
from autogen_ext.models.openai import OpenAIChatCompletionClient

server_params = SseServerParams(
    url="https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
)

async with McpWorkbench(server_params) as workbench:
    agent = AssistantAgent(
        name="Haystack (deepset Cloud)_assistant",
        model_client=OpenAIChatCompletionClient(model="gpt-4o"),
        workbench=workbench,
    )

    result = await agent.run("List recent Haystack (deepset Cloud) data")
    print(result.messages[-1].content)

Get your connection token →

Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by deepset Cloud. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.

Why Choose Vinkius

Vinkius connects your tools to AI with real-time monitoring and automatic cost savings — all from one dashboard.

Connect Haystack (deepset Cloud) now

Real-time monitoring

Live

visibility into every interaction

Connect your favorite tools to your AI and see exactly what's happening — every request, every response, in real time.

Built-in savings

60%

lower AI costs

Vinkius compresses data between your apps and your AI automatically. Lower bills every month — no configuration required.

Single dashboard

One

place for every integration

Every tool your AI connects to, managed from a single screen. One account, complete control.

Common questions about Haystack (deepset Cloud) MCP in AutoGen

You assign the tools to different agents. For example, give `run_pipeline` to an `AssistantAgent` and `get_pipeline` to a `UserProxyAgent` that critiques the results. They then collaborate through conversation to solve the problem.

Absolutely. One agent can perform an action with `run_pipeline`, and a second agent can use `get_pipeline` to check its configuration or `search_documents` to find alternative evidence. This adversarial setup is a core strength of AutoGen.

When you initialize your `AssistantAgent`, you pass a list of functions to its `llm_config`. You can create separate lists of tools from this MCP server and give one list to your 'executor' agent and a different list to your 'critic' agent.

An 'Architect' agent could use `list_pipelines` to review all available pipelines in a workspace. It could then recommend the best one for a specific task to an 'Executor' agent, kicking off the workflow.

The server itself is stateless and processes tool calls via your secure Vinkius endpoint. The primary consideration is the conversation history between your agents. Data passed between them, like query text for `search_documents` or pipeline IDs, will be in that transcript, which you control locally.

Use it with your favorite AI tools

Connect this server to Cursor, Claude, VS Code, and more.

OpenAI Agents SDK sdk-python

Google ADK sdk-python

Pydantic AI sdk-python

Vercel AI SDK sdk-typescript