How to Use the Elasticsearch Vector MCP in OpenAI Agents SDK

Q: How does my OpenAI Agent perform a kNN search with Elasticsearch Vector?

Your agent calls the search tool with a query vector and the target index name. The server runs the k-NN query inside Elasticsearch and returns the top matching document IDs and their scores. The OpenAI Agents SDK then passes these results back to your agent for the next step.

Q: What's the best way to handle index creation for Elasticsearch Vector with the OpenAI Agents SDK?

Have your agent first call listindexes to see if the required index exists. If not, it should then use createindex, passing the correct mapping for your densevector field. This programmatic setup avoids runtime errors and makes your agent more resilient.

Q: Can I use the OpenAI Agents SDK to manage document lifecycle in Elasticsearch Vector?

Yes, that's what it's for. Your agent uses indexdocument to add or update documents with their vectors. To remove one, it just calls deletedocument with the document's ID.

Run production-grade vector search on Elasticsearch with your OpenAI Agent, backed by built-in safety guardrails.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Elasticsearch Vector MCP to OpenAI Agents SDK

Create your Vinkius account to connect Elasticsearch Vector to OpenAI Agents SDK 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 Elasticsearch Vector with OpenAI Agents SDK

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Manage Elasticsearch Indexes From Your Agent

Your agent can now directly manage vector search indexes. Use `create_index` to set up a new `dense_vector` index with specific mappings and dimensions, right from your Python code. No more manual setup in Kibana. Before you index, your agent can check what's already there with `list_indexes` or get the details of a specific one using `get_index`. This gives your agent the context it needs to avoid errors and manage its own data environment. It's a solid way to build autonomous systems that handle their own setup.

Index and Search with the OpenAI Agents SDK

The `index_document` tool lets your agent add new documents with their corresponding vectors to any index. Your code prepares the data, and the agent handles the transaction with Elasticsearch. You can also remove specific entries using `delete_document` by its ID. The main event is the `search` tool. Your agent feeds it a query vector, and it runs a k-NN search to find the most similar documents. Since you're using the OpenAI Agents SDK, you can build guardrails around this, ensuring the agent only searches approved indexes or validates results before acting on them. This MCP Server makes it happen.

Build Self-Contained Search Applications

Connect this MCP Server and your agent can handle the entire lifecycle of a vector search task. It can check for an index, create it if missing, populate it with documents, and then run searches against it. This isn't just about running queries. It's about building agents that can maintain their own knowledge bases inside Elasticsearch. The agent becomes responsible for the data it uses, which is a big step up from just being a stateless tool-caller.

Setup guide

Set up Elasticsearch Vector MCP in OpenAI Agents SDK

Prerequisites

Python 3.10+ installed
openai-agents package (pip install openai-agents)
Active Vinkius subscription with a valid endpoint token

1

Install the SDK

Run pip install openai-agents to install the OpenAI Agents SDK. The MCP integration is built-in — no extra dependencies needed.
2

Connect via SSE transport

Use MCPServerSse with your Vinkius endpoint URL. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com. The SDK auto-discovers all Elasticsearch Vector tools at runtime.
3

Create your Agent

Pass the MCP to Agent(mcp_servers=[server]). The agent receives Elasticsearch Vector tools as native definitions — JSON schemas resolve automatically.
4

Run the agent

Call Runner.run(agent, prompt) to execute. The agent invokes the appropriate Elasticsearch Vector tools and returns structured results. Copy the full example on the right to get started.

agent.py

import asyncio
from agents import Agent, Runner
from agents.mcp import MCPServerSse

async def main():
    async with MCPServerSse(
        url="https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
    ) as server:
        agent = Agent(
            name="Elasticsearch Vector Agent",
            instructions="You have access to Elasticsearch Vector tools.",
            mcp_servers=[server],
        )
        result = await Runner.run(agent, "List recent transactions")
        print(result.final_output)

asyncio.run(main())

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 Elasticsearch Vector. 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 Elasticsearch Vector 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 Elasticsearch Vector MCP in OpenAI Agents SDK

Your agent calls the `search` tool with a query vector and the target index name. The server runs the k-NN query inside Elasticsearch and returns the top matching document IDs and their scores. The OpenAI Agents SDK then passes these results back to your agent for the next step.

Have your agent first call `list_indexes` to see if the required index exists. If not, it should then use `create_index`, passing the correct mapping for your `dense_vector` field. This programmatic setup avoids runtime errors and makes your agent more resilient.

Yes, that's what it's for. Your agent uses `index_document` to add or update documents with their vectors. To remove one, it just calls `delete_document` with the document's ID.

The OpenAI Agents SDK itself doesn't filter tools from a connected MCP server. You'd implement permission logic in your agent's code, creating guardrails that check the tool name before the agent is allowed to execute the call.

Your document content and dense vectors are passed through Vinkius's ephemeral infrastructure to your Elasticsearch instance. We don't store your data. The connection is secured, and your Vinkius token authenticates the requests, so only your authorized agent can interact with your indexes.

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