Ultra Context
The context API for AI agents
Same context. Everywhere.
Start on Claude Code. Continue on Codex.
Open source, realtime and invisible context infrastructure for the ones shipping at inference speed.
Documentation · API Reference · Changelog
What Claude Code knows, Codex doesn't. What your teammate is shipping right now? Your agent has no idea.
UltraContext captures every agent's context in realtime and makes it available to all of them. It's like having a personal context engineer everywhere. Continue a session in a different agent, or just ask what's happeming.
For example:
- "Codex, grab the last plan Claude Code made and implement it."
- "What's the team building today?"
- "What is Alex working on in Codex right now?"
Open source. Framework-agnostic. Customizable via the git-like Context API.
Features
| CLI | Auto-ingest Claude Code, Codex, and OpenClaw sessions with a terminal dashboard. |
|---|---|
| MCP Server | Share context everywhere. Built into the API, or run standalone via stdio. |
| Context API | Git-like context engineering API. Store, version, and retrieve agent context with zero complexity. |
How it works
-
Start sync. It captures all your agents' context in realtime.
-
Add the MCP server. Any agent gets full awareness of every other agent.
-
That's it. Ask questions, continue sessions, fork — your context is everywhere.
Install
Requires Node >= 22.
npm install -g ultracontext
Quick Start
ultracontext # start sync (daemon + dashboard)
That's it. UltraContext watches your agents, ingests context in realtime, and the dashboard shows everything.
ultracontext sync # start sync (daemon + dashboard)
ultracontext stop # stop daemon
ultracontext config # run setup wizard
ultracontext update # update CLI globally
Context API
For builders who want to go deeper. Git-like primitives for context engineering.
- Five methods — Create, get, append, update, delete. That's it.
- Automatic versioning — Every change creates a new version. Full history out of the box.
- Time-travel — Jump to any point in your context history.
- Framework-agnostic — Works with any LLM framework. No vendor lock-in.
Use the API standalone to build your own agents, or extend existing ones in UltraContext.
| SDK | Install | Source |
|---|---|---|
| JavaScript/TypeScript | npm install ultracontext | apps/js-sdk |
| Python | pip install ultracontext | apps/python-sdk |
JavaScript/TypeScript
npm install ultracontext
import { UltraContext } from 'ultracontext';
const uc = new UltraContext({ apiKey: 'uc_live_...' });
const ctx = await uc.create();
await uc.append(ctx.id, { role: 'user', content: 'Hello!' });
// use with any LLM framework
const response = await generateText({ model, messages: ctx.data });
Python
pip install ultracontext
from ultracontext import UltraContext
uc = UltraContext(api_key="uc_live_...")
ctx = uc.create()
uc.append(ctx["id"], {"role": "user", "content": "Hello!"})
# use with any LLM framework
response = generate_text(model=model, messages=uc.get(ctx["id"])["data"])
📚 Context API Guides
Store & Retrieve · Edit Contexts · Fork & Clone · View History
Star History
Documentation
- Quickstart — Get running in 2 minutes
- Guides — Practical patterns for common use cases
- API Reference — Full endpoint documentation
相關伺服器
Alpha Vantage MCP Server
贊助Access financial market data: realtime & historical stock, ETF, options, forex, crypto, commodities, fundamentals, technical indicators, & more
MCP Server Toolkit
A comprehensive toolkit for developing, testing, and deploying Model Context Protocol (MCP) servers.
YellowMCP
Reliability intelligence for remote MCP servers. Agent-native discovery for 1,700+ servers with uptime monitoring and latency benchmarks.
MCP Quickstart
A basic MCP server from the Quickstart Guide, adapted for OpenAI's Chat Completions API.
Valyu
Access Valyu's knowledge retrieval and feedback APIs.
E2B
Run code in secure sandboxes hosted by E2B
FAL FLUX.1 Kontext [Max]
A frontier image generation and editing model with advanced text rendering and contextual understanding, powered by the FAL AI API.
Linear Regression MCP
Train a Linear Regression model by uploading a CSV dataset file, demonstrating an end-to-end machine learning workflow.
Cupertino
Apple Documentation MCP Server - Search Apple docs, Swift Evolution, and sample code
Flame MCP Server
Provides live, up-to-date documentation for the Flame game engine.
Gradle Class Finder MCP
Find and decompile classes within Gradle dependencies.