Veta
Turns saved X posts and long-form articles into a private, searchable context base accessible via MCP. Lets coding agents import, search, read, and enrich the library using the user's existing browser session.
README
Veta
Turn saved X posts and long-form articles into a private, searchable context base. Veta runs on your computer, uses the browser already controlled by your coding agent, and exposes the library through MCP. There is no X API, OAuth app, API key, or extra browser extension.
Install it by asking your agent
Open a task in Codex, Claude Code, or Cursor and paste this:
Install Veta from https://github.com/nerdconf/veta by following
AGENT_INSTALL.md. Use your current browser to open my X bookmarks, ask me to sign in only if the session is closed, sync everything, and verify the result with a question about my library. Do not ask me to install or configure a browser integration.
The agent handles the clone, local setup, MCP registration, app startup, browser scrolling, import, and verification. The user only signs in to X when their existing browser session needs it.
What the workflow looks like
- Tell the agent to install and sync Veta.
- If X shows a login screen, sign in in that same agent-controlled tab.
- Let the agent scan and import bookmarks in resumable batches.
- Open the library at
http://127.0.0.1:4318or ask the agent questions directly. In future tasks the agent uses Veta's MCP tools automatically.
MCP is the interoperability layer, not the scraper itself. It gives Codex, Claude Code, and Cursor the same local tools to import, search, read, and enrich the database. Each product keeps using its own browser capability and its own existing subscription.
Privacy model
- The database and app state stay inside the local Veta installation.
- Veta has no analytics, hosted account, X credentials, or token store.
- X authentication remains inside the user's existing browser session.
- The active agent provider receives the saved text it processes or uses in an answer. “Local” describes storage and orchestration; it does not make the subscribed AI model offline.
.wrangler/and.veta/contain local state and are excluded from Git.
See PRIVACY.md for the complete boundary.
Legal
- Privacy Policy — English and Spanish
- MIT License
Contributor setup
npm install
npm run dev:local
The app uses http://127.0.0.1:4318; the optional in-app Codex question bridge
uses http://127.0.0.1:4317.
Useful checks:
npm test
npm run veta -- status
node scripts/veta-mcp-server.mjs
Veta is an early local-first release for macOS and Linux with Node.js 22.13+ and an installed Codex, Claude Code, or Cursor environment.
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.