Context Book

Created By
Aditya17 days ago
Stop re-explaining yourself to agents. Give it the right context right when it is needed. LLMs are stateless — each conversation starts from scratch. ContextBook gives your AI tools a persistent, searchable knowledge library to draw the right context when they need it. No bloatware, no pre-loaded junk. Just the right information, at the right time.
Overview

ContextBook

smithery badge License: MIT

Stop re-explaining yourself to Agents. Give it the right context, right when needed.

LLMs are stateless — each conversation starts from scratch. ContextBook gives your AI tools a persistent, searchable knowledge library to draw the right context when they need it. No bloatware, no pre-loaded junk. Just the right information, at the right time.

How It Works

flowchart LR
    AI["🤖 AI Clients\nClaude · Cursor · Windsurf"]
    Browser["🌐 Browser Dashboard"]
    MCP["🔌 MCP Server :8081\n8 Bearer-authenticated tools"]
    API["⚙️ REST API :8080\nOAuth 2.0 · Books · Pages · Search"]
    DB["🗄️ PostgreSQL\npgvector · pg_trgm"]
    VOYAGE["🧠 Voyage AI\nvoyage-4 (1024-dim)"]

    AI -- "MCP · Bearer Token" --> MCP
    Browser -- "Session · HTTP JSON" --> API
    MCP --- DB
    API --- DB
    API -- "OAuth 2.0 PKCE" --> AI
    DB --- VOYAGE

Two Go binaries share a PostgreSQL database:

  • API server (cmd/api) — the control plane: user login, OAuth 2.0, dashboard, book/page CRUD
  • MCP server (cmd/mcp) — the data plane: 8 MCP tools for AI agents, protected by Bearer tokens

MCP Tools

All tools require a valid Bearer token and are scoped to the authenticated user.

ToolDescription
book_create_or_updateCreate a Book or update its metadata
book_listPaginated list of Book metadata
book_getRetrieve all pages of a Book
page_insertPush an atomic page into a Book; embeds immediately
page_updateReplace a page's content; re-embeds
page_deleteRemove a page (indices not re-numbered)
page_searchSemantic search across all Books
readmeReturns the usage guide (call once per session)

Quick Start

Prerequisites

1. Set up the database

CREATE DATABASE contextbook_db;
\c contextbook_db
CREATE EXTENSION IF NOT EXISTS vector;
CREATE EXTENSION IF NOT EXISTS pg_trgm;

Migrations run automatically on API server startup.

2. Configure environment

cp .env.example backend/.env
# Edit backend/.env — set DATABASE_URL, API_KEY_SALT, VOYAGE_API_KEY

3. Run the backend

cd backend
go run ./cmd/api/main.go    # API + dashboard (:8080)
go run ./cmd/mcp/main.go     # MCP server (:8081)

4. Run the frontend (optional)

cd frontend
npm install && npm run dev    # Vite dev server on :5173

5. Connect an AI client

For Cursor (.cursor/mcp.json):

{
  "mcpServers": {
    "contextbook": {
      "url": "http://localhost:8081/mcp"
    }
  }
}

For any MCP-compatible client, point the server URL to http://localhost:8081/mcp.

Documentation

  • Backend README — API routes, auth flows, database schema, MCP tools, configuration
  • Frontend README — components, routing, design system, development setup
  • Architecture — full system overview with Mermaid diagrams

Project Structure

context-book/
├── backend/
│   ├── cmd/api/main.go          REST API + auth server
│   ├── cmd/mcp/main.go          MCP tool server
│   └── internal/
│       ├── api/                  REST handlers + routes
│       ├── auth/                 OAuth 2.0 PKCE, sessions, SSO
│       ├── context/              Book/Page business logic
│       ├── db/                  pgx queries + migrations
│       ├── embedding/            Voyage AI client
│       ├── logger/               slog + HTTP access logging
│       └── mcp/                  8 MCP tool handlers
├── frontend/                     React 19 + Vite + TypeScript SPA
│   └── src/
│       ├── App.tsx               Router + app shell
│       ├── lib/api.ts            HTTP client
│       └── components/           UI components
├── Dockerfile                    API server container
├── Dockerfile.mcp                MCP server container
└── go.work                       Go workspace

Contributing

  1. Fork the repository
  2. Create a feature branch (git checkout -b feature/my-feature)
  3. Make your changes
  4. Ensure the backend compiles (cd backend && go build ./cmd/api ./cmd/mcp)
  5. Ensure the frontend builds (cd frontend && npm run build)
  6. Commit and push
  7. Open a Pull Request

License

MIT — see the LICENSE file for details.

Server Config

{
  "mcpServers": {
    "context-book": {
      "url": "https://context-book-mcp-production.up.railway.app"
    }
  }
}
Project Info
Created At
17 days ago
Updated At
15 days ago
Author Name
Aditya
Star
-
Language
-
License
-
Category
Tags

Recommend Servers

View All
Bring your real authenticated browser session to AI coding agents. Local-first MCP server + Chrome MV3 extension. No cloud. No telemetry.
@Cubenest

peek records the user's actual logged-in browser (DOM via rrweb, console events, network metadata, optional response bodies via opt-in Deep capture) through a Chrome MV3 extension. The extension ships events through a native-messaging stdio bridge to a local MCP server (peek-mcp), which persists them to a SQLite database at ~/.peek/sessions.db. AI coding agents (Claude Code, Cursor, Cline, Windsurf) read sessions from the database via 10 MCP tools: Tool What it does list_recent_sessions List recently recorded sessions (id, origin, ts, event count). get_session_summary LLM-readable narrative summary of a session. get_session_console_errors Console errors recorded in a session. get_session_network_errors Failed/notable network requests in a session. get_user_action_before_error Last N user actions before a console error. generate_playwright_repro Generate a runnable Playwright test from a session. get_dom_snapshot Reconstruct the DOM at a given timestamp. query_dom_history Timeline of attribute/text changes for a selector. request_authorization Side-panel consent for write actions (Level 3). execute_action Dispatch a UI action (gated by permission level + destructive blocklist). Why local-first matters Every other "browser session for AI" tool ships to a vendor cloud. peek's SQLite + extension live on the user's machine — no remote endpoints, no telemetry. The privacy policy (docs/peek/PRIVACY_POLICY.md) is the source of truth. Install # 1. Add the MCP server to Claude Code claude mcp add peek -- npx -y @peekdev/mcp # 2. Install the Chrome extension from the Chrome Web Store # (link added once the CWS listing is approved)

9 hours ago
Alloy

2 days ago