# Danube AI

A gateway connecting coding assistants to external services and workflows via a single HTTP endpoint. It handles tool execution, skill storage, and spending limits.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI
- Author: danubeai
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/danubeai/danube-mcp
- Updated: 2026-03-01
- HTML: https://affpapa.org/en/nevibecode/mcp/danube-ai

## What it does

This MCP server runs over streamable HTTP at `https://mcp.danubeai.com/mcp`. It routes model calls into an ecosystem of 100+ services and 1400+ tools through unified authentication. Instead of configuring separate integrations for every external API, the assistant connects to a single remote bridge.

The server exposes over 30 tools. You can discover capabilities with `search_tools`, execute multiple calls at once using `batch_execute_tools`, or store multi-step routines with `create_workflow`. Financial controls are built directly into the schema: `get_wallet_balance` and `update_spending_limits` keep autonomous spending within defined bounds.

A valid Danube API key is mandatory. Without the `danube-api-key` header configured in your client, requests fail immediately. Developers can also interact with the platform directly using standalone Python and TypeScript SDKs.

## Requirements
- Needs: Python and uv (commands run via uvx)
- Remote server: `https://mcp.danubeai.com/mcp`

## Install

### Claude Code (native platform)

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

### Codex

**Platform CLI: remote server**
```bash
codex mcp add danube-ai --url https://mcp.danubeai.com/mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "danube-ai": {
            "url": "https://mcp.danubeai.com/mcp"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "danube-ai": {
            "type": "http",
            "url": "https://mcp.danubeai.com/mcp"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http danube-ai https://mcp.danubeai.com/mcp
```

## When to use it
- run external integrations through a single authenticated gateway
- bundle multiple tool calls into batch jobs using `batch_execute_tools`
- restrict agent budgets with `update_spending_limits`
- register and trigger reusable multi-step workflows

## Example prompts
- Check my Danube wallet balance and list current spending caps.
- Search for available data enrichment tools using `search_tools`.
- Create a new automated pipeline to handle incoming support tickets.

## Similar
- [workflows-create](https://affpapa.org/en/nevibecode/skills/workflows-create.md) — skill · 514 installs
- [outputai](https://affpapa.org/en/nevibecode/plugins/outputai.md) — plugin · ★ 446
- [inngest-steps](https://affpapa.org/en/nevibecode/skills/inngest-steps.md) — skill · 3800 installs
- [mastra](https://affpapa.org/en/nevibecode/skills/mastra.md) — skill · 83000 installs
- [workflow-guide](https://affpapa.org/en/nevibecode/skills/workflow-guide.md) — skill · ★ 222
- [voltagent-best-practices](https://affpapa.org/en/nevibecode/skills/voltagent-best-practices.md) — skill · 720 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