mcp server
langchain
LangChain integration for Ceki — real Chrome sessions for AI agents (TS + Python packages).
Set up
Runs on your computer. Pick your agent and operating system.
This project does not publish an install method we can turn into a config. Follow the README instead.
README
synced Oct 23, 2026Ceki LangChain integrations
LangChain integrations for Ceki — drive a real Chrome browser from your LangChain or LangGraph agent.
This monorepo ships two packages:
| Package | Lang | Path | Status |
|---|---|---|---|
@ceki/langchain-ceki | TypeScript / JS | packages/ts | v0.1.0 (pre-publish) |
langchain-ceki | Python | packages/python | v0.1.0 (pre-publish) |
Both packages are a thin wrapper over ceki-sdk / @ceki/sdk. They expose Ceki as a toolkit of structural LangChain tools — ceki_rent_browser, ceki_navigate, ceki_click, ceki_type, ceki_scroll, ceki_screenshot, ceki_snapshot, ceki_chat_send, ceki_stop — and let the agent's own LLM plan the sequence. No server-side natural-language endpoint, no LLM inside the wrapper.
Install
# TypeScript / Node
npm install @ceki/langchain-ceki @langchain/core
# Python
pip install langchain-ceki
Use
from langchain_ceki import CekiToolkit
toolkit = CekiToolkit(default_rent={"schedule_id": 4242})
tools = toolkit.get_tools()
# pass `tools` to any LangChain agent (create_tool_calling_agent, ...).
# When the run finishes:
await toolkit.aclose()
import { CekiToolkit } from "@ceki/langchain-ceki";
const toolkit = new CekiToolkit({ defaultRent: { scheduleId: 4242 } });
const tools = await toolkit.getTools();
// pass `tools` to any LangChain agent.
await toolkit.close();
Get an API key at ceki.me.
Use responsibly
Use only on sites you own or have authorization to operate on — your own apps, your own dashboards, public data within site Terms of Service, accessibility audits you're responsible for.
Related
- Ceki marketplace
ceki-sdkon PyPI and@ceki/sdkon npm — low-level CLI + SDK
License
MIT.