Nanobot

Why deploy nanobot on Render?

Nanobot is a lightweight, self-hostable AI agent with a built-in browser chat UI. It solves the problem of running your own private AI assistant with persistent memory, tool usage (web search, filesystem, shell, MCP servers), and session history without relying on third-party hosted services.

This template gives you a production-ready nanobot agent with persistent disk storage for session memory, secure token-based WebUI access, and all environment variable wiring already configured—no Docker setup or config file juggling required. One click deploys the gateway and chat interface together, with Render's Blueprint handling the service definition so you skip the manual infrastructure work. You get a live agent URL in minutes instead of debugging port bindings and persistence layers yourself.

Architecture

What you can build

After deploying, you'll have a private AI agent running on Render with a browser-based chat interface, persistent memory across sessions, and access to tools like web search, shell commands, and filesystem operations. The agent is protected by a secret token you control and stores conversation history on a persistent disk, so you can pick up where you left off. You'll need to bring your own Anthropic API key, and costs run about $7/month for hosting plus whatever you spend on Claude API calls.

Key features

• Built-in tool suite: Ships with web search/fetch, filesystem access, shell commands, and MCP server support enabled out of the box.
• Persistent session memory: Stores chat history and agent memory on a 1GB persistent disk so conversations survive restarts.
• Token-gated WebUI: Includes a browser-based chat interface secured by a configurable secret token for access control.
• Hardened demo mode: Optional public demo configuration with rate limiting, session isolation, SSRF protection, and restricted tools to safely expose the agent.
• One-click Render Blueprint: Deploys via render.yaml with pre-configured environment variables, non-root container, and workspace restrictions.

Use cases

• Indie developer deploys private AI assistant with persistent memory for daily tasks
• Startup founder hosts secure AI agent with web search for competitive research
• DevOps engineer runs self-hosted chatbot that executes shell commands remotely
• Team lead sets up locked-down demo agent for clients to preview capabilities

What's included

Prerequisites

• Anthropic API Key: Powers the AI agent's language model calls to Claude for generating responses.
• Nanobot Web Token: A secret access token you generate to protect your agent's web interface from unauthorized access.

Next steps

1. Open your service's onrender.com URL and paste your NANOBOT_WEB_TOKEN when prompted — You should see the chat interface load with an empty conversation ready for input
2. Send a test message like 'What can you help me with?' — You should receive a response from the agent within a few seconds, confirming the Anthropic API key is working
3. Test the web search tool by asking 'Search for today's tech news and summarize' — You should see the agent fetch live results and return a summary, confirming tools are enabled

Resources

• nanobot official site