manifest-mcp
Enables AI assistants to read and update a user's personal manifest (portable context) through a secure, token-based connector.
README
manifest-mcp
The server behind the link option at readthemanifest.net — one personal connector URL that lets Claude and ChatGPT read your manifest (your portable context) and save new facts to it.
This repo exists so you don't have to take our word for what the server does. Read it. It's three small files.
How it treats your manifest
Your connector URL looks like https://readthemanifest.net/mcp/<token>. The token is a
long random secret, and it does two jobs:
- It's the address. The server hashes the token (SHA-256) and uses the hash to label your storage slot. The token itself is never written down — not in a database, not in a log we keep.
- It's the key. Your manifest is encrypted (AES-256-GCM) with a key derived from the token (HKDF). So what sits in storage is a locked file the server cannot open on its own. It is only unlocked in the moment your AI asks for it, because the request carries the key.
No accounts, no email, no passwords. The honest consequence: lose the URL and the manifest is unrecoverable — there is nothing to reset. Export a copy anytime.
What "open source" does and doesn't prove
Reading this code tells you what the published server does. It cannot prove what a remote server is running — no open-source project's deployment can. If that distinction matters to you, Manifest's GitHub options keep your context entirely in your own hands, and this repo lets you run your own copy of this server (below) so the deployment is yours too.
The files
netlify/functions/mcp.mjs— the MCP endpoint (POST /mcp/:token). Stateless JSON-RPC over HTTP. Two tools:get_manifest(read) andadd_to_manifest(append one dated fact).netlify/functions/manifest-vault.mjs— vault management (POST /api/manifest-vault): create, replace, export, delete.netlify/functions/_lib/vault.mjs— the storage and crypto described above. Storage is Netlify Blobs.
Limits: manifest ≤ 150KB, note ≤ 4KB.
Run your own
Deploy this repo to any Netlify account (free tier works; Blobs is enabled automatically),
then add your own https://<your-site>/mcp/<token> as a custom connector. To try it
locally:
npm install
npx netlify dev --offline
# create a vault
curl -s -X POST http://localhost:8888/api/manifest-vault \
-H 'Content-Type: application/json' \
-d '{"action":"create","manifest":"# Manifest\n\nHello."}'
# read it the way your AI would (use the returned token)
curl -s -X POST http://localhost:8888/mcp/<token> \
-H 'Content-Type: application/json' \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"get_manifest","arguments":{}}}'
(Offline local dev stores vaults as files under .netlify/local-blobs/; deployed, it's
Netlify Blobs with strong read-after-write consistency.)
Usage counting
The server tallies how often its endpoints get used — an anonymous counter per
event name (vault_create, mcp_read, …) plus a per-day variant, in a separate
blob store. That's the whole record: no IPs, no user agents, no tokens, no
manifest content. See netlify/functions/_lib/counters.mjs — it's ~30 lines.
License
MIT.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.