ultracontext 1.1.3

The context API for AI agents

Context infrastructure for AI agents.

Documentation · API Reference · Changelog

---

All agents. One context.

Auto-capture and share your agents' context everywhere. Realtime. Open source.

Everyone is shipping with agents. Few are shipping with agents together.

Multiple people, multiple agents, multiple machines. Our contexts are spread everywhere. There's no standard for context engineering. No infrastructure to build on. No fundamental building blocks to agree on. So we decided to make it.

UltraContext is the context infrastructure. The API gives you git-like primitives for context engineering. The Hub lets you auto-capture, share, and collaborate across agents in realtime.

Install

Requires Node >= 22.

npm install -g ultracontext

The Hub

All agents. One context.

The Hub lets you auto-capture, share, and collaborate across agents in realtime.

Features

• Auto-capture — Ingests your agents' context in realtime. Zero config.
• Switch between agents — Pick up where one agent left off with another.
• Collaborate — Share contexts across your team. See what everyone sees. Realtime.
• Fork & clone — Continue contexts while preserving the full history.
• Own your data — Open source. Your contexts. Your rules.

How it works

1. A daemon runs in the background, watching your agents.
2. Contexts are ingested in realtime.
3. Your dashboard gets updated.

Quick Start

ultracontext          # start daemon + open dashboard
ultracontext config   # run setup wizard
ultracontext start    # start daemon only
ultracontext stop     # stop daemon
ultracontext status   # check if daemon is running
ultracontext tui      # open dashboard only

The default ultracontext command does everything: checks the daemon, starts it if needed, and opens the dashboard.

When you open an existing session, it forks the context — the original is always preserved and automatically versioned. A local caching layer prevents duplicate context creations and appends.

Add your own agents and extend behavior with the Context API. (Docs here)

The API

Context engineering built like Git.

The API gives you git-like primitives for context engineering, without the complexity.

Features

• 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.

The simplest way to control what your agents see. Replace messages, compact long context, replay decisions and roll back mistakes — all with a single API call.

Use the API standalone to build your own agents, or to 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