# Agentic-Ops Control Plane

> Agentic-Ops Control Plane is an open-source MCP server for developer tools. Hosted MLOps and SRE control plane for investigating drift and failing pods with agents. It can run locally on your machine or as a hosted remote server.

- **Kind:** MCP server
- **Pricing:** Open source
- **Platforms:** API, CLI, self-hosted, Web
- **Website:** https://agentic-ops-5ebv.onrender.com
- **Repository:** https://github.com/nolll77/agentic-ops
- **MCP Registry name:** io.github.nolll77/agentic-ops
- **Categories:** developer tools, cloud and infrastructure
- **Listing updated:** 2026-11-15

## About

This is a personal project, and the README says so: an exploration of how LLM agents can look into operational signals such as data drift, model staleness and OOMKilled pods. Its design leans on SRE-style controls rather than agent freedom, including deterministic policies, fail-closed authorization, asynchronous human review and cryptographic audit trails. The author states these are work in progress and that production qualification and EU AI Act compliance are not claimed.

It is published on PyPI for stdio use and as multi-architecture container images, and a hosted instance on Render exposes an MCP endpoint along with a web console, chat page and Swagger docs. The hosted status page shows an OpenRouter-backed model hub using free-tier models. Protected console and LLM-probe actions need a private ADMIN_API_KEY.

Release 0.2.0 adds server-sent events at /mcp/events with replay, simulated live telemetry streams for drift and latency, server-initiated sampling requests and an HMAC-sealed audit-pack resource that is marked as not third-party verifiable.

## Features

- **Drift and staleness triage:** Frames data drift, model staleness and OOMKilled pods as signals for agent-assisted investigation.
- **Fail-closed authorization:** Uses deterministic policies and human review before actions proceed.
- **Tamper-evident audit:** Keeps cryptographic evidence, including Merkle proofs and an HMAC-sealed audit pack.
- **Live telemetry streams:** Offers drift and latency resources over HTTP plus SSE, with simulated data by default.
- **Two transports:** Runs over stdio from PyPI or as a hosted streamable-http endpoint.
- **Web console and chat:** Includes browser pages, Swagger docs and a Slack interactive webhook.

## Good for

- Explore agent-assisted triage of ML drift alerts
- Prototype governed approval flows for operational actions
- Study audit-trail designs for agent platforms
- Try MCP sampling and SSE event streaming against a live endpoint

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http agentic-ops https://agentic-ops-5ebv.onrender.com/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://agentic-ops-5ebv.onrender.com/mcp
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "agentic-ops": {
      "url": "https://agentic-ops-5ebv.onrender.com/mcp"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "agentic-ops": {
      "type": "http",
      "url": "https://agentic-ops-5ebv.onrender.com/mcp"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "agentic-ops": {
      "serverUrl": "https://agentic-ops-5ebv.onrender.com/mcp"
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "agentic-ops": {
      "type": "http",
      "url": "https://agentic-ops-5ebv.onrender.com/mcp"
    }
  }
}
```

## FAQ

### Is it ready for production?

No. The README says production qualification is not claimed and the features are engineering work in progress.

### How can I run it?

Install the agentic-ops package from PyPI for stdio, or pull the container image from GHCR or Docker Hub; a hosted endpoint is also listed.

### Does the telemetry reflect real systems?

The live drift stream defaults to a simulated source, which the README marks honestly.

### Does it claim EU AI Act compliance?

No. The audit pack is marked as not third-party verifiable and not legal certification.

### Is Agentic-Ops Control Plane free?

Yes. Agentic-Ops Control Plane is open source, so you can use and self-host it for free.

### How do I add Agentic-Ops Control Plane to Claude Code?

Run this command in your terminal: claude mcp add --transport http agentic-ops https://agentic-ops-5ebv.onrender.com/mcp

### What are the best alternatives to Agentic-Ops Control Plane?

The closest alternatives to Agentic-Ops Control Plane are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

## Alternatives

- [REA](https://ai-powered.com/mcp/rea): REA is an open-source MCP server for developer tools.
- [Codebase Memory](https://ai-powered.com/mcp/codebase-memory): Codebase Memory is an open-source MCP server for knowledge and memory.
- [GitHub](https://ai-powered.com/mcp/github-2): GitHub is an open-source MCP server for developer tools.
- [Serena MCP: the IDE for your agent](https://ai-powered.com/mcp/serena-mcp-the-ide-for-your-agent): Serena MCP: the IDE for your agent is an open-source MCP server for developer tools.
- [XERJ](https://ai-powered.com/mcp/xerj): XERJ is an open-source MCP server for knowledge and memory.
- [Ouroboros](https://ai-powered.com/mcp/ouroboros): Ouroboros is an open-source MCP server for developer tools.
- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [Concord MCP](https://ai-powered.com/mcp/concord-mcp): Concord MCP is a freemium MCP server for developer tools.

---
Source: https://ai-powered.com/mcp/agentic-ops-control-plane
