Start with one goal. Keep long-running agent work verifiable.
Tell the Agent you already use what you want done. LoopX keeps the goal, work,
needed decisions and verifiable results durable behind the scenes, so the next
session can continue without making you learn LoopX internals first.
Work with Codex, Claude Code, DeepSeek Harness and other runtimes. Advanced docs still expose the local-first architecture, ownership model and research workbench.
Get started · Why LoopX · See it in action · Evidence · Contribute · Docs · 简体中文
Start with one goal
1. Install LoopX once. Requires Python 3.11+ and
Node.js 22.22.3+ (24 LTS recommended).
python3 -m pip install --upgrade loopx && loopx workflow-skills --install
Install options and upgrades
2. Tell your Agent the outcome. Restart your Agent host, open your project,
then send one natural-language goal:
$loopx speed up this project's slowest workflow; compare approaches and verify the gains
LoopX turns that into a visible plan, decisions that need you, and verification
as needed. Claude Code uses /loopx <complex task>; see
host setup for other runtimes. Prefer a guided first
run? Choose a task below.
📰 News
- 2026-10-25 · 🤝 Long-horizon agent research: collaborators welcome — Exploring breakthroughs in long-horizon outcomes under equal budgets, through feedback, replanning and reproducible evaluation on EdgeBench and real-world tasks. Academic labs and industry teams are welcome to collaborate, comparing outcomes and total cost. Discuss collaboration.
- 2026-10-20 · 🚀 LoopX 2.0 — A shared TypeScript state layer and long-horizon research workbench. Migration guide · Engineering story.
- 2026-10-14 · 🤖 Agentic software — Connecting goals, evidence and human feedback to engineering delivery.
- 2026-10-02 · 🧠 Long-horizon bootstrapping & RSI — Four months of developing LoopX with LoopX: public project history and a design perspective on system-level self-improvement.
- 2026-09-18 · 🧪 LoopX × LHTB — A 46-task, same-model study provides empirical support for trading more compute for better outcomes: LoopX 1.0.3 Heartbeat continued runs reach mean reward 17.3% above Plain Codex and 10.6% above native Goal.
- 2026-09-06 · 🎉 LoopX 1.0 — A Personal Agent Workspace to follow agent work, act on decisions, and inspect delivered results across sessions. Workspace guide.
- 2026-08-02 · ⏱️ 200+ hour trajectories shared on X — Early owner-run examples spanning 200+ elapsed hours drew broad attention on X, showing work across execution, waits, human decisions and recovery. Trajectories and limits.
All releases
Why LoopX
You should not have to learn LoopX internals before the first useful action.
The default path is goal-first: describe the outcome, let the Agent prepare the
plan, and inspect the machinery only when a decision or debugging need calls
for it.
An agent can finish a task in one session. Long-running work is harder:
objectives change, owner decisions appear, evidence goes stale, and agents
hand work to peers. A timer can keep a loop running without making it useful.
- Continue without starting over. Keep the Goal, accepted requirements,
decisions and evidence across sessions and restarts.
- Coordinate without dropping context. Connect delegated work, dependent
artifacts, review and returned results while keeping responsibility explicit.
- Judge progress by what changed. Inspect deliverables, supporting and
refuting evidence, costs and the next decision—not just activity.
This is the beginning of the README. Read the rest at the original. Read the full README ↗
