LoopX

**Give your agents a goal. Keep the work moving.** The open, local-first control plane for long-horizon agents and personal agent teams.
_{Keep goals, decisions and evidence across sessions. Work with Codex, Claude Code, DeepSeek Harness and other supported runtimes.} [![License](https://img.shields.io/badge/license-Apache--2.0-blue.svg)](LICENSE) [![Release](https://img.shields.io/github/v/release/loopx-project/loopx?filter=v*&display_name=tag)](https://github.com/loopx-project/loopx/releases/latest) [![GitHub contributors](https://img.shields.io/github/contributors-anon/loopx-project/loopx?color=blue)](https://github.com/loopx-project/loopx/graphs/contributors) [![Discord](https://img.shields.io/badge/Discord-Join-5865F2?logo=discord&logoColor=white)](https://discord.gg/XmGgQyCFZd) [![TypeScript core](https://img.shields.io/badge/core-TypeScript-3178C6?logo=typescript&logoColor=white)](docs/architecture/rfcs/typescript-control-plane-migration-v0.md) [![Python](https://img.shields.io/badge/python-3.11%2B-blue.svg)](pyproject.toml) [![Local first](https://img.shields.io/badge/control--plane-local--first-brightgreen.svg)](docs/public-private-boundary.md) [![Loop Agents](https://img.shields.io/badge/status-loop%20agents%20supported-brightgreen.svg)](docs/product/release-readiness.md)

[Get started](#try-loopx) · [Workspace](#meet-the-personal-agent-workspace) · [LHTB results](#lhtb-results) · [Docs](https://loopx-project.github.io/loopx/docs/) · [简体中文](README.zh-CN.md)

LoopX Personal Agent Workspace: a Work map showing task dependencies, responsible agents, progress and decisions waiting for the owner

_{Source-built App on an authored Workspace stories replay; no live Agent execution · watch the 32-second walkthrough} **Start in Codex App in two steps.** Install once: ```bash python3 -m pip install --upgrade loopx && loopx workflow-skills --install ``` Restart Codex App, then type this in a thread opened on your project: ```text $loopx fix the open PR review feedback and keep the patch reviewable ``` _{LoopX connects the project, plans the work as Todos and sets up the heartbeat that keeps it moving. Claude Code uses /loopx <complex task> · other Agent hosts} **[LHTB](https://zli12321.github.io/LHTB/index.html) · 46 tasks · GPT-5.6 Sol:** LoopX 1.0.3 Heartbeat reaches **0.4948 mean reward** — **+17.3% vs Plain Codex**, **+10.6% vs native Codex Goal**.
_{Results and pass rates ↓}

--- **More verified work. Less human attention.** LoopX gives agents durable goals, bounded continuation, peer ownership and recoverable handoffs. Your runtime provides the model and tools; LoopX keeps track of what to do next, what is accepted, and when to ask you. | What you want to do | Start here | | --- | --- | | Keep a coding or research agent working across sessions | [Install and start](#try-loopx) | | Manage personal projects, schedules and decisions in one place | [Personal Agent Workspace](#meet-the-personal-agent-workspace) | | Let agents collaborate and deliver verifiable results | [Agent collaboration guide](docs/product/use-cases/cross-runtime/README.md) | ## Meet the Personal Agent Workspace Keep long-horizon goals in one local-first workspace. Goals, attention, conversations, tasks, files, schedules, and recovery stay durable across days, restarts, and harnesses. Reopen a project, inspect the previous turn’s state and evidence, and continue the next permitted action. LoopX 1.0 brings these long-horizon control states into the Personal Workspace. It gives you one place to: - see what needs you, what is running, what is being watched, and what is scheduled or stopped; - configure Goal capabilities, distinguish machine defaults from Goal overrides, and preview changes before applying them; - steer a live turn, queue a message, or use the async inbox from a connected Lark conversation with explicit Goal/Agent/session routing; - inspect deliverable files, periodic reports, and their supporting evidence; - continue across Codex, Claude Code, direct-model, and other registered Agent sessions without losing Goal state or evidence; - review protected changes through typed preview, explicit confirmation, and receipts while LoopX state—not the browser—remains authoritative. For Manager group conversations, LoopX keeps message visibility separate from Turn authority; see the bilingual [Lark Manager context and authority contract](docs/reference/protocols/lark-manager-context-authority-v0.md). ```bash loopx dashboard ``` `loopx dashboard` is the supported browser/PWA launch path. You can also download native desktop previews from the [1.0 release](https://github.com/loopx-project/loopx/releases/tag/v1.0.0); they reuse the same loopback services and Goal state. Apple Silicon macOS supports signed App updates that pair the shell with its bundled runtime, plus repair and recovery. Python 3.11+ is required; the App is ad-hoc signed, not notarized. Windows preview installers currently use manual updates and a separately installed CLI. [Desktop installation, updates, and source development](apps/desktop/loopx-control-plane/README.md).

Capability settings and reproducible workspace scenarios

Real Workspace recording: configure child-task capacity and allowed responsibility domains

From a source checkout, run `python -m demo.workspace serve` to explore a community event, a home-energy comparison, and a neighborhood website release. Each has four work roles, 18 tasks, two decisions, and two watches. The screenshot above comes from this reproducible workspace. [Scenarios and replay instructions](demo/workspace/README.md).

[Watch the full 32-second walkthrough](docs/assets/personal-workspace/loopx-dashboard-launch.mp4) · [Read the workspace guide](docs/guides/personal-workspace-user-guide.md) · [Try the five-minute tour](docs/guides/personal-workspace-trial-guide.md) ## Why LoopX An agent can finish a task in one session. Long-running work is harder: objectives change, owner decisions appear, evidence goes stale, agents hand work to peers, and a scheduler can keep spending after no useful transition remains. Chat memory and a timer are not enough to govern that. LoopX keeps the durable control state in one compact layer:

Your harness runs the turn. LoopX carries the work.

Agent runtimes execute the work. LoopX governs the state that lets engineering, research, discovery, and operations loops continue across runs. It is not another agent framework or a provider-specific orchestration runtime.

One turn is a governed state transition—not just a timer.

A useful mental model is an **[agent-native Kanban for long-running work](docs/development/control-plane-course/00-concept-primer.md)**. Cards carry identity, authority, evidence, and continuation. Moves are validated operators such as claim, gate, monitor, and writeback. The board is a projection; LoopX state remains the source of truth. Registered agents are peers. Claims, leases, task boundaries, capabilities, and typed continuation decide who acts next; no durable leader identity is required. LoopX is useful when you run: - multi-day engineering, research, benchmark, or experiment objectives; - issue and PR loops that must preserve scope, evidence, and review state; - recurring heartbeat or monitor work; - projects with owner, safety, publication, or private-data gates; - peer-agent teams where ownership, leases, and handoff matter; - creator, research, or operations workflows whose progress must remain legible to a non-engineering operator. LoopX is not an autonomous production controller. Dangerous permissions, publishing, production writes, and final ownership stay with the human. ### Personal Agents, Teams, and Self-Improving Workflows [Meta Muse](https://about.fb.com/news/2026/09/introducing-muse-personal-ai-agent/) and [Grok Bot](https://x.ai/news/introducing-grok-bot) make persistent personal agents and delegated work a familiar product idea. LoopX approaches that space as an open, provider-neutral control plane for agents you already run—not as a hosted replacement for either product. - **Personal agent:** use the shipped Workspace and connected Lark surfaces to inspect goals, steer work and resolve decisions. The [persistent steward and semantic handoff RFC](docs/architecture/rfcs/capable-manager-semantic-handoff-v0.md) extends this toward one capable front door for a team; the complete journey remains under qualification. - **Agent team:** share goals without erasing worker ownership. Claims, evidence and explicit returns keep implementation and review connected. See the [three-agent collaboration demo](examples/collaboration-delivery/README.md) for a bounded example, including corrections and two review rounds. - **Self-improving workflows:** connect feedback to the next attempt through [Reward Memory](loopx/capabilities/reward_memory/README.md), and evaluate candidate changes through [Explore](loopx/capabilities/explore/README.md). Both are optional. The link to **recursive self-improvement (RSI)** is the engineering loop—propose, test, review, retain—not a claim that LoopX has demonstrated autonomous model improvement or compounding capability gains. A personal steward is an interaction role, not a second source of authority. Background execution still needs an available host and configured runtime. The [overall roadmap](docs/architecture/rfcs/loopx-overall-roadmap-v0.md) separates shipped foundations from the next end-to-end acceptance milestones. ## Evidence ### LHTB: Higher Mean Reward With the Same Model **LoopX 1.0.3 Heartbeat reaches 0.4948 mean reward across 46 matched tasks:** **+17.3% over Plain Codex** and **+10.6% over native Codex Goal**. All three use GPT-5.6 Sol, max reasoning effort, and disabled Web Search. [Long-Horizon Terminal-Bench](https://zli12321.github.io/LHTB/index.html) goes beyond coding: these tasks span research reproduction, scientific simulation, multimodal analysis, professional workflows, games, and systems work. | Execution mode | Mean reward ↑ | Strict pass rate (≥0.95) | Pass rate (≥0.80) | | --- | ---: | ---: | ---: | | Plain Codex | 0.4218 | 7/46 (15.2%) | 12/46 (26.1%) | | Native Codex Goal | 0.4475 | 4/46 (8.7%) | 14/46 (30.4%) | | **LoopX 1.0.3 Heartbeat** | **0.4948** | **7/46 (15.2%)** | **15/46 (32.6%)** | The ≥0.80 threshold is a supplementary, post-hoc view; ≥0.95 remains the benchmark's strict solved threshold. No score in this study equals 0.80, so these counts also match the study brief's >0.80 view. Against native Goal, task-level rewards improve on **23**, tie on **13**, and regress on **10** tasks. Strict solves match Plain Codex, so the mean gain is not a claim that every task improves or that more tasks are fully solved than Plain. Mean reward also credits partial progress. One effective trial per task and mode; designated replacement trials and unequal runtime/budgets, including longer budgets for some Heartbeat replacements. These are observed system results, not an equal-budget efficiency result or an isolated causal estimate of LoopX's effect. [Interactive results, task comparisons and methodology](https://loopx-project.github.io/loopx/benchmarks/lhtb/) · [Five-arm study and limits](benchmark/LHTB/studies/five-arm-gpt56sol-max/README.md) · [Public task-level scores](benchmark/LHTB/studies/five-arm-gpt56sol-max/data.json) Beyond benchmarks, LoopX also has inspectable long-running project evidence. The public OpenViking contribution sequence and the redacted, owner-run Auto ML showcase each span **200+ hours of elapsed loop lifetime**, preserving bounded turns, decisions, and evidence updates. This measures wall-clock project time, not continuous model execution or unattended production autonomy. Open each visual to inspect the public-safe task graph, evidence branches, and cross-turn decisions; each case states its source and reproducibility boundary.

Earlier 200+ hour project trajectories: OpenViking and ML experiments

### Open-Source Issue Fix **200+ hour public contribution arc: PR delivery and reusable fix knowledge evolve together.**

Open-source issue-fix trajectory linking focused PR delivery with reusable LoopX capabilities

LoopX's creator uses this path as an [OpenViking contributor](https://github.com/volcengine/OpenViking/pulls?q=is%3Apr+author%3Ahuangruiteng). The represented public contribution sequence spans more than 200 elapsed hours from its first PR creation to the latest represented review or update. The [Issue-Fix capability](loopx/capabilities/issue_fix/README.md) keeps rolling repository context, revision-stamped fix knowledge, and reviewer-facing preferences separate; linked PRs plus current checkout source and tests remain authoritative. ### Auto ML Experiment **Redacted owner-run showcase: a 200+ hour experiment arc keeps hypotheses, matched evidence, invalid lineages, running replicates, and promote/stop gates visible in one graph.**

Auto ML Experiment trajectory with experiment lineages, evidence gates, and promotion decisions

The redacted public-safe graph preserves decision lineage across that 200+ hour elapsed window. It is an owner-run showcase, not a claim of continuous compute, independent reproduction, or a production result.

### Used In Real Projects - **Independent user · `>13h` C++ accuracy run.** The user reported that a multi-stage task stayed aligned, triggered public research, adopted a [public code-memory tool](https://github.com/DeusData/codebase-memory-mcp), and improved final precision. [Read the evidence boundary](docs/showcases/cases/independent-cpp-accuracy-long-run.md). - **Independent user · `4d` unattended run.** The user reported four days without human intervention, useful ongoing work, and a periodic report surface. [Read the redacted case](docs/showcases/cases/independent-four-day-unattended-agent.md). - **Independent user · `7` merged PRs.** A LoopX-attributed Engine refactor is visible in a [public issue](https://github.com/zilliztech/mfs/issues/166) and seven merged PRs; attribution and the reported `1B+` token scale remain user reports. [Inspect the case](docs/showcases/cases/independent-public-engine-refactor.md). - **Independent user · vLLM optimization merged upstream.** [Zheng Cai](https://x.com/zigzagcai/status/2109952174822408443) reports using LoopX as the workflow engine for profiling, research, experiments, and correctness checks. [vLLM PR #54951](https://github.com/vllm-project/vllm/pull/54951) merged a sparse-indexer optimization; its reported **32.59%** total-token throughput gain is for GLM-5.3-Flash at 1M input tokens, 4×H200, TP4, C1. [Inspect the workflow, full conditions, and evidence boundary](docs/showcases/cases/independent-vllm-long-context-optimization.md). These are four featured independent-user cases, not the full inventory. Browse the [complete Showcase catalog](docs/showcases/README.md) for contributor cases, creator dogfooding, reproducible demos, and explicit evidence-strength labels. ### Exploratory Benchmark Studies - **[SWE-Marathon](https://loopx-project.github.io/loopx/benchmarks/swe-marathon/):** Five execution modes on 15 matched tasks compare self-verification, scores, and cost. More self-verification did not consistently yield higher scores. - **[LHTB × LoopX](https://loopx-project.github.io/loopx/benchmarks/lhtb/):** The [results above](#lhtb-results) lead into five execution mechanisms, task-level comparisons, regressions, and exploratory task-type analysis. - **[DeepSWE behavior analysis](https://loopx-project.github.io/loopx/benchmarks/deepswe/behavior-discovery/)** (Chinese): Selected cases examine how domain hints relate to requirement retention and verification choices, offering mechanism hypotheses for further testing. SWE-Marathon and LHTB have one effective trial per task and mode; DeepSWE uses selected cases and post-hoc analysis. More inspectable surfaces: - the [public homepage](https://loopx-project.github.io/loopx/) for the product narrative, quick start, and long-running evidence; - the [complete Showcase catalog](docs/showcases/README.md) and its [bilingual hosted index](docs/showcases/index.html); - the [Agent collaboration guide](docs/product/use-cases/cross-runtime/README.md) for dependent artifacts, independent review and returned results; - the public [user manual](https://my.feishu.cn/wiki/CaL5wMk9ui17ngkWzeUcMlAYnZg). ## Try LoopX Install once: ```bash python3 -m pip install --upgrade loopx && loopx workflow-skills --install ``` ### Start From Your Agent Restart your Agent host, then start from a session opened on your project: | Host | Type | | --- | --- | | Codex App, Codex CLI | `$loopx ` | | Claude Code | `/loopx ` | | DeepSeek Harness | Install the [native DSH plugin](packages/dsh-loopx-plugin/README.md), select the `loopx` skill, then describe the task | LoopX reuses existing project state or connects the project, plans the work as Todos and sets up the host's loop. `loopx status` shows the goal, any decision waiting for you and the next Todo; `loopx doctor` diagnoses the install. Requires Python 3.11+ and Node.js 22.22.3+ (24 LTS recommended). [Installing LoopX](docs/guides/installing-loopx.md) covers native Windows, `pipx`, upgrades, rollback and uninstall.

Other hosts, loop drivers and custom runners

| Host | Recommended start | Loop driver | | --- | --- | --- | | Codex App | `$loopx `, or choose `loopx` from `/skills`, in a thread opened on the project. | Codex App heartbeat automation, refreshed from `quota should-run.scheduler_hint` | | Codex App over SSH | `loopx agent-onboard --agent-type codex-app-ssh --project .` | The returned visible `/goal ` | | Codex CLI | Start `codex` in the project, then `$loopx ` or `/skills`. | Visible `/goal `; no hidden headless execution by default | | Claude Code | Install the opt-in adapter, then run `/loopx ` followed by `/loop`. | Native Claude Code `/loop` gated by LoopX | | KunlunCode | Run `loopx-kunluncode connect --project . --goal-id --agent-id `, add a bounded todo, then run `loopx-kunluncode run --project .`. | Native Goal Pro through app-server; LoopX writes completion and quota only after strict verification | | OpenCode | Install the static command facade; opt in to `--with-goal-bridge` for recurring goals. | OpenCode command facade and explicit goal bridge | | Pi | Install the opt-in goal extension with `loopx slash-commands --install --surface pi`, then use `/loopx ` from a trusted Pi session. | Visible Pi goal extension gated by LoopX quota (`loopx_goal_activate` + `agent_settled` continuation) | | ZCode | Install the skill facade with `loopx slash-commands --install --surface zcode`, then invoke the `$loopx` skill (or `/loopx `) from a ZCode session in the project. | The ZCode session's own turn loop; every continuation enters through `quota should-run` | | Antigravity CLI (agy) | Install the skill facade with `loopx slash-commands --install --surface agy`, then invoke the `loopx` skill (or `/loopx `) from an `agy` session in the project. | The session's native `/goal` loop (audited until ``) with `schedule` self-wakes while the session lives; the facade instructs every turn/wake to re-enter through `quota should-run` — advisory pacing, not a host-enforced gate | | Kiro CLI | Install the skill facade with `loopx slash-commands --install --surface kiro-cli`, then run `/loopx ` from a `kiro-cli` session in the project. | The session's native `/goal --max Done when: ` loop, with the acceptance criteria stated inside the goal statement because the host derives them from it, bounded by the host's own iteration budget (default 5) and settled through the built-in `goal` completion contract; the facade instructs every turn and iteration to re-enter through `quota should-run` — advisory pacing, not a host-enforced gate | | DeepSeek Harness (dsh) | Install the [native DSH plugin](packages/dsh-loopx-plugin/README.md), select the `loopx` skill, and describe the task. The [dsh goal-mode adapter](loopx/dsh_goal_mode/README.md) remains available for headless turns. | Native same-session continuation and GoalBar, or headless dsh segments; both remain gated by LoopX authority | | Cursor, shell, or custom runner | `loopx start-goal --guided --project . --goal-text ""`, or call LoopX from your runner. | Your shell, scheduler, or runner | The exact, copy-ready setup messages and host recovery paths live in [Getting Started](docs/guides/getting-started.md). Host integrations can inspect the [Codex App host command registry contract](docs/reference/protocols/codex-app-host-command-registry-v0.md), the [Codex CLI packaged install path](docs/product/runtimes/codex-cli/codex-cli-packaged-install.md), the [Claude Code adapter](loopx/claude_goal_mode/README.md), the [KunlunCode native Goal adapter](loopx/kunluncode_goal_mode/README.md), the [Kiro CLI goal-mode adapter](loopx/kiro_cli_goal_mode/README.md), or the [DeepSeek Harness turn adapter](loopx/dsh_goal_mode/README.md). See the [60-second DSH × LoopX Replan recording and reproducible fixture](docs/showcases/cases/dsh-loopx-replan-demo.md) for the native path: install the plugin, select one skill, change a material constraint, and inspect the preserved decision trail. For custom runners, start with the [minimal custom runtime example](docs/guides/minimal-custom-runtime-example.md) (`python3 examples/custom-runtime-minimal-cli-turn-smoke.py`), then the full [Embed LoopX in Your Agent Runner](docs/guides/custom-agent-runner-integration.md) guide and the [worker bridge install contract](docs/integrations/worker-bridge-install-contract.md). The core tick is deliberately small: ```text loopx quota should-run # should this registered agent act now? loopx todo claim # who owns this slice? loopx todo update # what changed? loopx refresh-state # what should the next turn see? loopx quota spend-slot # account for a completed, validated slice ``` Clone-based install is only for contributors who want the live canary wrapper: ```bash git clone https://github.com/loopx-project/loopx ~/loopx ~/loopx/scripts/install-local.sh loopx doctor ```

### First-Run Feedback If LoopX works for you, a one-minute public issue helps us learn what a real first run looks like. It is optional, contains no telemetry, and should not include logs, paths, credentials, internal project names, or goal contents: - [First-run feedback](https://github.com/loopx-project/loopx/issues/new?template=first_run.yml) - [Usage story for longer runs](https://github.com/loopx-project/loopx/issues/new?template=usage_story.yml) `loopx first-run-report` prints the same prefilled link locally without sending anything. Basic usage statistics default on after first-use disclosure: a daily random-ID heartbeat for platform support and continued use, plus separate ID-free CLI counts. No content is collected. Disable both in Settings → Capability Center or with `loopx usage-ping disable` / `LOOPX_USAGE_PING=0`; inspect payloads with `loopx usage-ping status`. See [Basic usage statistics](docs/reference/usage-ping.md). ## Capabilities LoopX keeps the architecture explicit so the same governed outcome can survive a change of agent harness or external provider. The terms describe different boundaries rather than interchangeable kinds of plugin: | Boundary | Meaning | Go deeper | | --- | --- | --- | | **Kernel** | Owns durable goal, todo, gate, evidence, quota, recovery, and scheduling truth. | [Architecture](docs/architecture.md) | | **Capability** | Defines a stable, provider-neutral contract for producing one bounded, verifiable caller outcome from LoopX state. | [Capability catalog](loopx/capabilities/README.md) | | **Provider** | Calls an external system or local implementation and returns bounded observations, effect results, and readback. | [Provider responsibilities](docs/reference/extensions.md#runtime-responsibilities) | | **Extension** | Packages and operates an optional provider through explicit install, readiness, enable, upgrade, disable, and rollback lifecycle. | [Extension lifecycle](docs/reference/extensions.md#runtime-lifecycle) | Host declarations such as `--available-capability shell` describe observed execution support. They are runtime capacities in this product map, not product capabilities and not permission grants. The effective capability still applies its own policy and authority checks before proposing a transition. ### Core Control-Plane Promises The Kernel folds its mechanics into five questions. Each question delivers one product promise on top of any agent harness: objective → long-horizon state; next → semantic decisions; human judgment → human-agent collaboration; evidence → recovery; continuation → governance. | Question | What LoopX keeps visible | | --- | --- | | What is the objective? | The active goal, explicit scope, and current authority. | | What happens next? | Ordered user and agent todos, ownership, claims, and leases. | | What needs human judgment? | Concrete user gates instead of a vague "waiting for owner." | | What evidence changed? | Compact run history, validation, blockers, and accepted writeback. | | May the loop continue? | Quota, capabilities, safe fallback, scheduler hints, and stop conditions. | ### Control-Plane Surface | Surface | What it does | Start with | | --- | --- | --- | | Goal state and status | Tracks active state, todos, claims, gates, evidence, run history, and first-screen attention. | `loopx status`, `loopx diagnose`, `loopx review-packet` | | Quota and interaction contract | Decides whether a turn should deliver, ask, wait, self-repair, or stay quiet. | `loopx quota should-run`, [quota allocation](docs/quota-allocation.md) | | Agent runtime bridges | Keeps Codex App, Codex CLI, Claude Code, and generic workers aligned with the same guard. | `loopx heartbeat-prompt`, `loopx codex-cli-bootstrap-message`, `loopx worker-bridge` | | Operator surfaces | Renders compact status without making the browser the state authority. | `loopx serve-status`, [dashboard](apps/presentation/dashboard/README.md) | | Session dash | Starts a live single-page panel that tracks fleet progress: sessions, their goals, and each goal's status/todo progress, with result statistics; auto-refreshes in place. | `loopx dash`, [session dash design](docs/product/surfaces/session-dash-panel-design.md) | | External projections | Projects todos and gates into collaboration surfaces while LoopX remains authoritative. | `loopx lark-kanban`, [Lark Kanban adapter](docs/integrations/lark-kanban-control-plane-adapter.md) | | Domain capabilities | Packages repeatable work lanes such as issue fixing, content operations, value connector planning, ML experiment advice, benchmark evidence, and Explore. | `loopx issue-fix`, `loopx content-ops`, `loopx value-connectors`, `loopx ml-experiment`, `loopx benchmark`, [Explore](loopx/capabilities/explore/README.md) | | Experimental context learning | Lets named registered agents trial provider-neutral Reward Memory through ignored, default-off project configuration. OpenViking is one provider option, not a global dependency. | `loopx reward-memory experiment-status`, [Reward Memory architecture](loopx/capabilities/reward_memory/README.md) | | Governance patterns | Captures reusable routing, gate, evidence, projection, and planning shapes. | [interaction patterns](docs/concepts/interaction-pattern-catalog.md), [state model](docs/state-interaction-model.md) | The shipped primitives include lifetime goals, concrete user gates, audited safe fallbacks, peer todo ownership, quota and steering, compact run history, evidence-backed handoff, a read-first management surface, project-level value signals, and public/private boundary checks. ### Product Capability Paths Capabilities turn those generic primitives into outcome-owned work lanes. Start from the outcome, then inspect the current registered implementation and its write boundary: | You need to... | Capability | Start with | | --- | --- | --- | | Turn a public issue into a reviewable, evidence-backed change | [Issue Fix](loopx/capabilities/issue_fix/README.md) | `loopx capability show issue-fix --format json` | | Qualify the exact final diff before delivery | [Change Quality](loopx/capabilities/change_quality/README.md) | `loopx capability show change-quality-qualification --format json` | | Notice busy-but-off-goal work rounds before the periodic review | [Progress-Review Sentinel](loopx/capabilities/progress_review/README.md) | `loopx capability show progress-review-sentinel --format json` | | Preserve a changing stack of already reviewed branches | [Integration Branch](loopx/capabilities/integration_branch/README.md) | `loopx capability show integration-branch-reconcile --format json` | | Explore uncertain research without losing hypotheses and findings | [Explore](loopx/capabilities/explore/README.md) | `loopx capability show explore --format json` | | Rebase decisions on current evidence and verified outcomes | [Decision Context](loopx/capabilities/decision_context/README.md) | `loopx capability show decision-context --format json` | | Produce scheduled or progress-triggered reports with receipts | [Periodic Report](loopx/capabilities/periodic_report/README.md) | `loopx capability show periodic-report --format json` | Run `loopx capability list --format json` for the authoritative catalog in the installed release. A capability detail reports its user value, maturity, provider readiness, entry commands, write boundaries, protocols, and durable validation. Browse the [human-readable capability index](loopx/capabilities/README.md) to choose by outcome; use [Extensions and Capabilities](docs/reference/extensions.md) when installing or building a provider. ### Runtime Responsibilities | Role | Responsibility | | --- | --- | | **Agent** | Plans, analyzes, uses tools, and performs one bounded action through a host/runtime. | | **Provider** | Calls external systems and returns observations, effect results, and readback. | | **Capability** | Defines the caller outcome, normalizes provider output, validates it, and proposes a typed transition. | | **Kernel** | Owns durable todos, gates, monitors, accepted writeback, quota, recovery, and scheduling. | The execution path is `Agent -> Capability -> Provider`; the control path returns `Provider readback -> Capability transition -> Kernel`. An extension is how an optional provider is packaged and managed, not another control-plane owner. See [Architecture](docs/architecture.md) and [Extensions and Capabilities](docs/reference/extensions.md). ## Advanced Paths The first useful loop does not require every optional surface. Add these only when the work needs them. Inspect the current goal's read-only capability catalog before enabling an advanced path: ```bash loopx configure-goal --goal-id ``` Without `--execute`, this reports current/default state, fit, boundaries, and copyable commands without changing project state. ### Recurring Work Presets Safe presets cover daily triage, changelog drafts, and PR watching. Start with the [beginner preset guide](docs/product/foundations/beginner-loop-presets.md). ```bash loopx preset list loopx preset show daily-triage ``` Preset inspection is read-only. For a connected recurring goal, `loopx ready-score --goal-id --agent-id ` reports whether the loop is ready to run repeatedly. ### Governed Turns LoopX can generate one pure, bounded turn decision from a validated receipt, fresh quota state, and a provider-neutral budget. The current Codex CLI quickstart and activation contract are documented in [LoopX Turn for Codex CLI](docs/product/runtimes/codex-cli/loopx-turn-codex-cli-quickstart.md). ### Explore Graph and Harness Explore is supported, optional, and default-off. It works best when a task has a measurable offline evaluation, baseline, treatment, and guardrails; it is not a substitute for production approval. Start with the [Explore capability](loopx/capabilities/explore/README.md) and its [Lark presentation mapping](loopx/capabilities/explore/README.md#presentation-sink-lark-mapping). ### Review Agent Work Use `loopx review-packet` for a compact owner-facing view of decisions, evidence, validation, and unresolved gates. The [intelligent management surface](docs/product/surfaces/intelligent-management-surface.md) describes the operator model; the [project-level reward model](docs/product/foundations/project-level-reward-model.md) describes conservative value signals across output quantity, quality, token cost, and user attention cost. For a concrete peer workflow, start with the [Agent collaboration guide](docs/product/use-cases/cross-runtime/README.md): a builder consumes an analyst's artifact, obtains independent review, handles an owner correction and returns the checked result. The guide distinguishes the runnable same-host example from the earlier cross-runtime design sketch. ### App and Projection Paths - Local read-first UI: [dashboard guide](apps/presentation/dashboard/README.md) - Public product overview: [public homepage](https://loopx-project.github.io/loopx/) - Documentation portal: [hosted docs](https://loopx-project.github.io/loopx/docs/) - Feishu/Lark projection: [Lark Kanban adapter](docs/integrations/lark-kanban-control-plane-adapter.md) - Generic host integration: [integration guide](docs/integration.md) - Custom multi-agent runner: [minimal custom runtime example](docs/guides/minimal-custom-runtime-example.md), then [custom runner integration](docs/guides/custom-agent-runner-integration.md) Optional projections make state easier to inspect; they do not become the source of truth. ### Operating and Recovery Start daily inspection with: ```bash loopx status loopx history --goal-id your-project-goal loopx quota should-run --goal-id your-project-goal ``` Automatic turns must check quota first and append spend only after validated writeback. Quiet skips, preflight failures, and dry-run previews do not spend. When a user gate blocks one lane, a separately audited safe fallback may continue, but it must not bypass the gate. Peer agents use `loopx todo claim` before delivery and `loopx todo update` after validation so ownership and evidence remain visible. Scheduler cadence follows `quota should-run.scheduler_hint`; installed Codex App automations acknowledge the current hint through the returned `ack_hint.cli_args`. Collision recovery, monitor semantics, self-repair, and the exact operator commands are maintained in [Getting Started](docs/guides/getting-started.md), [Quota Allocation](docs/quota-allocation.md), and [Long-Task Cadence Policy](docs/operations/long-task-cadence-policy.md). Before publishing public docs or examples: ```bash loopx check \ --scan-path README.md \ --scan-path docs/ \ --scan-path examples/ ``` ## Current Technical Directions LoopX has three active strategic programs plus an architecture and research incubator. These are direction signals, not delivery promises; `main`, released artifacts, and stable reference contracts remain the source of shipped truth. - **Long-Horizon Benchmarks and Evidence:** reproducible capability evidence and controlled mechanism research across complementary benchmark environments. [Direction tracker](https://github.com/loopx-project/loopx/issues/3243) - **Operator Surface and IM Integration:** an operator workspace, session records, and bounded collaboration surfaces. The Personal Workspace is shipped; unified steward, execution, and recovery journeys remain under qualification, with `@maxliux5` as implementation lead. [Direction tracker](https://github.com/loopx-project/loopx/issues/3244) - **Shared Goal Authority and Cross-host Coordination:** provider-neutral coordination for explicitly shared goals, with NoKV as an unpromoted provider candidate rather than a new control-plane authority. [Direction tracker](https://github.com/loopx-project/loopx/issues/3245) - **Architecture and Research Incubator:** Effect Program hardening, TypeScript parity migration, hierarchical stride, research exploration, human attention, artifact lifecycle, and memory utility work at explicitly different maturity levels. [Direction tracker](https://github.com/loopx-project/loopx/issues/3246) Read the canonical [Technical Directions map](docs/project/technical-directions.md) for stages, promotion gates, contributor-safe cuts, and ownership boundaries. Use the pinned [GitHub Discussion](https://github.com/loopx-project/loopx/discussions/2851) for community discussion. Core control-plane reliability continues as the shared foundation beneath these programs. ## Advanced Documentation [Product website](https://loopx-project.github.io/loopx/) · [Blog](https://loopx-project.github.io/loopx/blog/) · [Developer Book](https://loopx-project.github.io/loopx/docs/book/en/) Start with the path that matches your current task. Use the hosted [documentation portal](https://loopx-project.github.io/loopx/docs/) for the published docs site; the [documentation index](docs/README.md) remains the complete source map. This list stays selective; each category index owns its deeper documents and versioned protocols. ### Use and Operate - [Getting Started](docs/guides/getting-started.md): install, connect, diagnose, daily workflow, heartbeats, dashboard, development, and commands. - [User Manual](https://my.feishu.cn/wiki/CaL5wMk9ui17ngkWzeUcMlAYnZg): public onboarding, concepts, FAQ, and selected cases. - [Operations](docs/operations/README.md): goal continuation, todo, cadence, attention, and authority workflows. - [Quota Allocation](docs/quota-allocation.md) and [Heartbeat Automation Prompt](docs/heartbeat-automation-prompt.md): scheduler eligibility, spend, and scheduled continuation. - [Dashboard](apps/presentation/dashboard/README.md) and [Status Data Contract](docs/status-data-contract.md): operator-facing state and projection contracts. - [Release Readiness](docs/product/release-readiness.md): install/update paths, compatibility gates, release notes, and safe-to-depend-on surfaces. ### Understand the Control Plane - [Architecture](docs/architecture.md): lifetime-goal invariant and kernel. - [State Interaction Model](docs/state-interaction-model.md): actors, stores, interaction contract, and writeback. - [Concepts](docs/concepts/README.md): reusable routing, gate, evidence, projection, and planning patterns. - [Product Foundations](docs/product/foundations/README.md): Loop Engineering principles, project-level reward, and reward-style replanning. - [Product Vision](docs/product/vision.md): the broader Loop Agent direction. ### Integrate and Extend - [Integration Guide](docs/integration.md) - [Minimal Custom Runtime Example](docs/guides/minimal-custom-runtime-example.md) - [Custom Agent Runner Integration](docs/guides/custom-agent-runner-integration.md) - [Integrations](docs/integrations/README.md): runtime, host, collaboration, and external-system adapters, including worker bridge and Lark. - [Extensions and Capabilities](docs/reference/extensions.md) ### Build and Review LoopX - [Developer Guide](docs/development/README.md): contributor workflows, benchmark development, documentation layout, and quality gates. - [Reference and Protocols](docs/reference/README.md): stable contracts and versioned implementation protocols, including host command and reward memory architecture. - [Control-Plane Developer Course](docs/development/control-plane-course/README.md): nine Chinese, code-led lectures. - [Testing and Quality](docs/development/testing-and-quality.md): validation layers and risk-based checks. - [Public/Private Boundary](docs/public-private-boundary.md): safe fixtures, examples, evidence, and publication. ### Inspect Outcomes - [Showcase Catalog](docs/showcases/README.md): public-safe cases and evidence labels. - [Research and Evidence](docs/research/README.md): benchmark investigations and source-backed findings. - [Update Notes](docs/update-notes/README.md): public-safe progress notes. ### Project and Community - [Current Technical Directions](docs/project/technical-directions.md) - [Project Governance](.github/GOVERNANCE.md) - [Contributing](CONTRIBUTING.md) and [Contributor Tasks](docs/development/contributor-tasks.md) - [Authors and Contributors](docs/project/authors.md) - [Project History](docs/project/history.md) - [Name and Marks](docs/project/trademarks.md) - [Brand Guide for External Use](docs/project/brand-guide.md): how projects and companies may reference LoopX, its integrations, and its artwork - [ADOPTERS](ADOPTERS.md): voluntary self-attested adoption directory - [Ecosystem Adoption](docs/community/ecosystem-adoption.md) - integrations, sampling, and derivatives we observe and track ## Partner Projects LoopX welcomes collaboration with other open-source projects to build the long-running agent ecosystem. Our confirmed partners include: - [OpenViking](https://github.com/volcengine/OpenViking) - Self-evolving context database for AI agents - [NoKV](https://github.com/NoKV-Lab/NoKV) - AI native distributed file system ## Community and Feedback LoopX is already running real long-running agent goals and is under active development. The most useful feedback comes from real long-running agent projects: where the control plane helped, where it felt heavy, and which gates or handoffs disappeared from view. - Use [GitHub Issues](https://github.com/loopx-project/loopx/issues) for reproducible bugs, install problems, and feature requests. - Open PRs for docs fixes, showcase writeups, and small public-safe examples. - Join the [Discord community](https://discord.gg/XmGgQyCFZd), or use Lark or WeChat below. See [Support](.github/SUPPORT.md) for channel routing, service boundaries, and official publication sources.

_{Lark: scan to join directly
WeChat: huangrt00 · mention LoopX in the friend request}

## Contributing External contributors should start with [Contributor Tasks](docs/development/contributor-tasks.md) for public, claimable work and [Contributing](CONTRIBUTING.md) for setup, validation, and boundary rules. Project roles and public history are recorded in [Governance](.github/GOVERNANCE.md), [Authors and Contributors](docs/project/authors.md), and [Project History](docs/project/history.md). LoopX keeps local active state separate from the public repository. Do not commit `.loopx/`, legacy `.codex/goals/`, live `ACTIVE_GOAL_STATE.md`, raw benchmark traces, credentials, private logs, or operator artifacts. ## Current Status LoopX 1.0 is a usable local control plane for long-running agent work and is entering broader adoption. It is not a full agent platform, an agent runtime, or an autonomous production controller. Today LoopX ships a durable state kernel for goals, typed todos and decision scopes, peer claims and leases, evidence and writeback, quota-aware scheduling, and cross-turn continuation. Guided start, recurring heartbeat, isolated Codex CLI turns, evidence-backed Issue-Fix admission, optional Explore and auto research paths, public validation canaries, and the multi-project Personal Workspace all build on that shared control state. Support levels remain explicit. The state and CLI contracts are the stable center; several host integrations and advanced paths are optional, default-off, or experimental. LoopX does not grant credentials, approve destructive or production actions, publish on a user's behalf without authorization, or turn an unverified run into evidence of success. Current investment is organized through the [Technical Directions map](docs/project/technical-directions.md): long-horizon benchmark evidence, operator surface and IM integration, shared-goal cross-host coordination, and an explicitly staged architecture and research incubator. ## Star History

_{Generated every six hours from GitHub's official stargazer timestamps using a repository-authorized workflow. A snapshot is published only when the fetched rows match GitHub's current star count; GitHub's image cache may delay refreshes.}

## License Copyright 2026 LoopX contributors. Created by [`@huangruiteng`](https://github.com/huangruiteng); see the public [creator and contributor record](docs/project/authors.md). Apache License 2.0 beginning with `v0.4.8`. See [LICENSE](LICENSE) and [NOTICE](NOTICE). Releases through `v0.4.7` remain under their original MIT terms; the historical text and notice are preserved in [LICENSE-MIT](LICENSE-MIT). The [licensing policy](docs/project/licensing.md) explains the version, contribution, patent-grant, and open-core boundaries. You must not use AI generation, code rewriting, cross-language translation, or project renaming as grounds to remove or evade copyright, attribution, or NOTICE retention requirements under the applicable license. If the distributed material is a derivative work of LoopX, those obligations still apply. This clarification does not restrict modifications or redistribution permitted by the applicable license. See [reuse and attribution](docs/project/licensing.md#reuse-ports-and-attribution) and [name and marks guidance](docs/project/trademarks.md); a derivative distribution must not be presented as an official LoopX release. [osai-verify: eb42dd9cf910399988f0]: #