# Neonia: The Cloud Backend for AI Agents

The Neonia gateway gives an agent remote tools: JSON filtering, tool discovery and cloud memory that survives between sessions.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 5
- Author: Neonia-io
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/Neonia-io/agent-mcp-examples/tree/HEAD/mcp-registry
- Updated: 2026-06-04
- HTML: https://affpapa.org/en/nevibecode/mcp/neonia-the-cloud-backend-for-ai-agents

## What it does

Neonia runs as an MCP gateway at `https://mcp.neonia.io/mcp`, reached over Streamable HTTP. Through it an agent gets `neonia_data_jq_filter`, which runs queries against large JSON on the server and hands back only the filtered value, plus `neonia_sys_memory_note` and `neonia_sys_memory_search` for storing and recalling rules and preferences. A separate tool, `neonia_sys_memory_lesson`, records findings in a cause-and-effect shape. The point is that a big file never lands in the context window.

The `Neonia-io/agent-mcp-examples` repository ships working examples in Python (LangGraph, SmolAgents), TypeScript (Vercel AI SDK) and Rust (Rig). Each folder covers one pattern: single-tool filtering, chained filters, auto-discovery, note memory, lesson memory. Setup steps live in the README of the folder you pick.

Two keys are required: `NEONIA_API_KEY` and `OPENROUTER_API_KEY`, both placed in a `.env` file inside the example directory. Without them the agent cannot reach the gateway, so set them up before the first run.

## Requirements
- Remote server: `https://mcp.neonia.io/mcp`

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport http mcp-registry https://mcp.neonia.io/mcp
```

### Codex

**Platform CLI: remote server**
```bash
codex mcp add mcp-registry --url https://mcp.neonia.io/mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "mcp-registry": {
            "url": "https://mcp.neonia.io/mcp"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "mcp-registry": {
            "type": "http",
            "url": "https://mcp.neonia.io/mcp"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http mcp-registry https://mcp.neonia.io/mcp
```

## When to use it
- query a 5MB JSON file without flooding the context window
- pull in a tool the agent was never given
- keep user preferences across separate sessions
- store an architectural lesson as cause and effect
- build an agent on LangGraph, SmolAgents, Vercel AI SDK or Rig

## Example prompts
- Connect to the Neonia gateway and filter `data.json` — return only the sum of the amount field
- Save a rule in Neonia memory: always answer in English
- Search Neonia memory for migration lessons and show them before we start

## Similar
- [REA](https://affpapa.org/en/nevibecode/mcp/rea.md) — MCP server · ★ 392262
- [GitHub](https://affpapa.org/en/nevibecode/mcp/github.md) — MCP server · ★ 33989
- [Netdata](https://affpapa.org/en/nevibecode/mcp/netdata.md) — MCP server · ★ 81163
- [browser-use](https://affpapa.org/en/nevibecode/mcp/browser-use.md) — MCP server · ★ 119231
- [Codebase Memory](https://affpapa.org/en/nevibecode/mcp/codebase-memory.md) — MCP server · ★ 47577
- [mcp-server](https://affpapa.org/en/nevibecode/mcp/mcp-server.md) — MCP server · ★ 43902

---
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