AMAP-ML/LongHorizon-Harness — The long-horizon computer-use harness.

Description and use

**Summary:**

AMAP-ML/LongHorizon-Harness is an open-source project that enables long-horizon AI agent execution across desktop apps and the command-line interface (CLI) for complex workflows. It preserves task state, ensures reliable progress, and integrates with popular AI tools like Claude Code, Codex, and OpenClaw. This harness helps automate repetitive manual operations and connect tools into repeatable flows.

**Value Proposition:**

The primary value of AMAP-ML/LongHorizon-Harness lies in its ability to remove repetitive manual operations from a workflow, making it a valuable tool for automating tasks and increasing productivity. By integrating with popular AI tools, it enables users to leverage advanced AI capabilities in their workflows, making it an attractive solution for businesses and developers looking to streamline their operations.

**Practical Adoption Path:**

To adopt AMAP-ML/LongHorizon-Harness, users can start by evaluating the project's implementation signals, such as its API, SDK, CLI, and language metadata. Next, they can explore the project's documentation and community resources to understand its features and capabilities. Once familiar with the project, users can begin integrating it into their workflows, starting with simple tasks and gradually moving on to more complex ones. As they become more comfortable

Problem

AMAP-ML/LongHorizon-Harness helps remove repetitive manual operations from a workflow.

Outcome

High for an OSS candidate: recent activity, adoption, and ecosystem signals are strong enough for a serious pilot.

Workflow

• remove manual work
• connect tools into repeatable flows
• schedule operational tasks

Evidence

The long-horizon computer-use harness. Run AI agents across desktop apps and the CLI for extended periods while preserving task state and making reliable progress on complex workflows. Features fresh-context execution, durable verified state, independent auditing, recoverable progress, and native Claude Code / Codex / OpenClaw integration.

Source