TinyDB Remote MCP
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@TinyDB Remote MCPInsert a new user with name Aarav Sharma and email aarav.sharma@example.com"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
TinyDB Remote MCP
A small, practical Model Context Protocol (MCP) server that stores data in TinyDB. It is designed as a learning project for people who want to understand how to build an MCP server and connect it to an MCP client.
The detailed, step-by-step project setup and deployment notes are in NOTES.md.
What this MCP does
The server exposes three tools:
insert_data— stores one JSON document in TinyDB.get_all_data— returns every stored document.delete_all_data— removes every stored document.
For example, insert_data accepts:
{
"data": {
"name": "Aarav Sharma",
"email": "aarav.sharma@example.com",
"role": "admin",
"active": true
}
}Related MCP server: tldraw-mcp
Run locally
This project uses uv to manage dependencies.
uv sync
uv run python tinydb_remote_server.pyThe local Streamable HTTP endpoint is:
http://127.0.0.1:10000/mcpTinyDB creates db.json at runtime. It is local runtime data and is not
intended to be committed to version control.
Use the MCP
MCP Inspector
Start the Inspector locally:
npx @modelcontextprotocol/inspectorSelect Streamable HTTP and enter either the local endpoint above or your deployed endpoint:
https://<render-service-name>.onrender.com/mcpYou can also list the tools from the command line:
npx @modelcontextprotocol/inspector --cli https://<render-service-name>.onrender.com/mcp --transport http --method tools/listOpenAI web app
Enable Developer mode in ChatGPT under Settings > Security and login.
Then, from Settings > Plugins (or https://chatgpt.com/plugins), create
an app using your deployed MCP URL:
https://<render-service-name>.onrender.com/mcpAfter it is created, select the app in a ChatGPT project and ask it to insert or retrieve test records.
Codex, Claude web app, and Claude Code
Add the deployed /mcp URL as a Streamable HTTP MCP server in the client’s
MCP or connector settings. Once connected, the client can call
insert_data, get_all_data, and delete_all_data.
Deployment and security notes
See NOTES.md for the full Render deployment guide. When creating
the Render service, add **/*.md as an Ignored Path under Advanced >
Build Filters so Markdown-only changes do not trigger a deployment.
This example server is public and unauthenticated. Use test data only, and do not store credentials, personal information, or other sensitive data. On Render’s Free service, the TinyDB data is ephemeral and may be lost after a restart or redeploy.
This server cannot be deployed
Maintenance
Related MCP Connectors
An MCP server that provides read access to your cloud storage providers, bank accounts and more.
Fillo MCP server — provision, scaffold, publish, and query forms from your coding agent.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceProvides long-term memory for chatbots through a TinyDB-backed MCP server. Enables storing, searching, and managing JSON records with schema validation without requiring external databases.1MIT
- FlicenseAqualityDmaintenanceMinimal MCP server for editing tldraw .tldr files via JSON manipulation. Headless, no browser needed.181-
- AlicenseNot gradedqualityDmaintenanceA local MCP-style server for managing task records and seeded documents through controlled tool functions, using SQLite and JSON seed data.MIT
- AlicenseAqualityAmaintenanceAn MCP server that returns random JSON things — people, words, values, coordinates, and an always-empty list.4MIT