How to Use the Neptune.ai (ML Experiment Tracking) MCP in OpenAI Agents SDK

Q: How do I search for a specific experiment using Neptune.ai (ML Experiment Tracking) and the OpenAI Agents SDK?

You just ask your agent. It will use the searchruns tool with a query matching your request. The SDK handles the API call, and you get back a structured list of runs that you can then ask the agent to inspect further with getattributes.

Q: Can my OpenAI agent compare two different ML runs?

Yes. The agent would use searchruns twice with different criteria, then call getattributes for each resulting run ID. You can then instruct the agent to summarize the differences in parameters or metrics between the two.

Connect your OpenAI Agent to Neptune.ai to query ML runs and model metadata with built-in safety checks.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Neptune.ai (ML Experiment Tracking) MCP to OpenAI Agents SDK

Create your Vinkius account to connect Neptune.ai (ML Experiment Tracking) 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 Neptune.ai (ML Experiment Tracking) with OpenAI Agents SDK

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Inspect ML Runs from Your Agent

Give your agent the ability to dig into your experiment history. It can use the `search_runs` tool to find specific training runs based on your criteria, then pull the details with `get_attributes`. You get direct answers about model performance without leaving your chat interface. This isn't just for prototypes. The OpenAI Agents SDK has guardrails that validate agent actions before they run. That means you can trust the agent to query your production ML projects, knowing every call is checked and logged automatically in your OpenAI dashboard.

Audit Models with Your OpenAI Agent

Your agent can get a full inventory of your ML assets. Start with `list_projects` to see what's available, then use `list_models` to get a catalog of trained models inside a specific project. It’s a fast way to check what's been registered and what's ready for deployment. Because every tool call is traced, you have a perfect audit trail. You can see exactly which projects and models your agent inspected and when. This Neptune.ai MCP Server gives you a programmatic way to enforce governance and track how your ML assets are being accessed.

Zero-Config Tool Discovery

Connecting your agent is straightforward. After a `pip install`, you point the `MCPServerStreamableHttp` client at the Vinkius endpoint URL. That's it. Your agent automatically discovers the available tools—`search_runs`, `get_user`, and the others—with no manual mapping required. For production agents that run frequently, set `cacheToolsList=True` in the constructor. This tells the agent to cache the tool definitions from this MCP connection, which cuts down on startup latency for subsequent runs. It’s a small tweak that makes a noticeable difference.

Setup guide

Set up Neptune.ai (ML Experiment Tracking) 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 Neptune.ai (ML Experiment Tracking) tools at runtime.
3

Create your Agent

Pass the MCP to Agent(mcp_servers=[server]). The agent receives Neptune.ai (ML Experiment Tracking) tools as native definitions — JSON schemas resolve automatically.
4

Run the agent

Call Runner.run(agent, prompt) to execute. The agent invokes the appropriate Neptune.ai (ML Experiment Tracking) 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="Neptune.ai (ML Experiment Tracking) Agent",
            instructions="You have access to Neptune.ai (ML Experiment Tracking) 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 Neptune.ai. 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 Neptune.ai (ML Experiment Tracking) 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 Neptune.ai (ML Experiment Tracking) MCP in OpenAI Agents SDK

You just ask your agent. It will use the `search_runs` tool with a query matching your request. The SDK handles the API call, and you get back a structured list of runs that you can then ask the agent to inspect further with `get_attributes`.

Yes. The agent would use `search_runs` twice with different criteria, then call `get_attributes` for each resulting run ID. You can then instruct the agent to summarize the differences in parameters or metrics between the two.

The SDK gives you built-in guardrails and agent-to-agent handoffs. You're not just running a script; you're building a system that can safely query your Neptune.ai projects. Every action is also auditable through the OpenAI platform's tracing.

Yes. Your token is stored as a secret on the Vinkius platform, which runs the MCP server. The agent only interacts with the Vinkius endpoint and never has direct access to your Neptune.ai credentials.

The server only reads your Neptune.ai experiment metadata. This includes things like run IDs, logged parameters, metrics, and project names. Vinkius sandboxes each server process, and connections are ephemeral, so your experiment data is only touched for the duration of the API call.

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