# Omnara

Connects Claude Code to the Omnara platform to manage and execute autonomous AI agents across sandboxes and external machines.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 3220
- Author: omnara-ai
- License: Apache-2.0 · policy=link-only · description only; the file stays with the author
- Source: https://github.com/omnara-ai/omnara
- Updated: 2026-10-24
- HTML: https://affpapa.org/en/nevibecode/mcp/omnara

## What it does

Omnara provides an environment for running persistent AI agents. This MCP server connects your coding assistant to the remote platform at `https://app.omnara.com/mcp` over HTTP. It allows agents to dispatch commands to isolated sandboxes, control execution environments, and route requests across different model providers.

Under the hood, the system uses background workers and a Postgres database that records every step of an agent's run. If an execution fails, the state remains intact in the event log. Running the self-hosted version requires Docker Compose, while the CLI demands Node.js 22 or newer. Without exposing a self-hosted instance to a public URL, cloud sandboxes will not be able to talk back to your deployment.

## Requirements
- Needs: Docker
- Remote server: `https://app.omnara.com/mcp`

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport http omnara https://app.omnara.com/mcp
```

### Codex

**Platform CLI: remote server**
```bash
codex mcp add omnara --url https://app.omnara.com/mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "omnara": {
            "url": "https://app.omnara.com/mcp"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "omnara": {
            "type": "http",
            "url": "https://app.omnara.com/mcp"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http omnara https://app.omnara.com/mcp
```

## When to use it
- run commands inside an isolated sandbox environment
- delegate background coding workflows to dedicated machines
- persist agent execution logs and state transitions in Postgres
- route agent tasks to external model providers

## Example prompts
- Launch an agent in Omnara using `agent.yaml` to check the operating system and free disk space.
- Stream recent event logs from the active Omnara agent ID.
- Dispatch a background command to the `default-pool` sandbox.

## Similar
- [deep-agents-orchestration](https://affpapa.org/en/nevibecode/skills/deep-agents-orchestration.md) — skill · 16100 installs
- [swarm](https://affpapa.org/en/nevibecode/skills/swarm.md) — skill · 8800 installs
- [design-agent](https://affpapa.org/en/nevibecode/skills/design-agent.md) — skill · 8300 installs
- [multi-agent-orchestration](https://affpapa.org/en/nevibecode/skills/multi-agent-orchestration.md) — skill · 2100 installs
- [ai-agents-architect](https://affpapa.org/en/nevibecode/skills/sickn33-ai-agents-architect.md) — skill · 1500 installs
- [open-multi-agent-orchestration](https://affpapa.org/en/nevibecode/skills/open-multi-agent-orchestration.md) — skill · 710 installs

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md