jkf87

jkf87 / openclaw-harness

Public

OpenClaw Plan→Work→Review workflow engine with model routing

12
4
80% credibility
Found Mar 30, 2026 at 12 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Shell
AI Summary

OpenClaw Harness orchestrates AI agents through plan-work-review cycles to build and refine software projects with automatic issue detection, progress notifications, and smart AI selection.

How It Works

1
🔍 Discover the Harness

You come across OpenClaw Harness, a smart organizer that gets AI helpers to team up on building and fixing projects just like you imagined.

2
📥 Set It Up

You add it to your AI workspace with a quick install, creating folders for everything to run smoothly.

3
🔗 Link Your AI Thinkers

You connect affordable AI services that handle Korean tasks best, so they can think and create efficiently.

4
💭 Describe Your Idea

You simply tell it what to do, like 'build a simple todo app' or 'fix this bug in my code'.

5
🚀 Watch the Magic Happen

AI planners break your idea into steps, workers build it together, and reviewers spot and fix any mismatches automatically.

6
📱 Stay Updated Live

You get fun notifications on your chat app showing progress, like '60% done' or 'fixing a small gap'.

🎉 Enjoy Your Project

Your perfectly built app or fixed code is ready, saving you hours and matching your vision exactly.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 12 to 12 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is openclaw-harness?

OpenClaw Harness is a shell-based workflow engine that orchestrates AI agents through a Plan-Work-Review cycle for coding tasks, automatically detecting gaps like scope creep or missing features and triggering one fix loop. It solves AI drift by routing tasks to optimal models—GLM for Korean NLP or low-cost work, GPT variants for complex code gen—while pushing real-time status to Telegram or Discord. Developers get structured outputs from user requests, with budget profiles capping daily spend.

Why is it gaining traction?

Its model routing engine picks cost-effective LLMs based on task complexity and language, slashing bills on routine work while reserving premium models for architecture or security. The gap detection from five review angles (DoD, security, perf) keeps agents aligned without constant intervention, and Korean optimization hooks non-English users. OpenClaw GitHub integration via clawhub install and sessions_spawn makes it a seamless skill for OpenClaw workflows.

Who should use this?

Korean backend devs prototyping APIs or refactoring, needing cheap GLM routing without English bias. Teams running parallel agent workers on mid-complexity tasks like auth modules or tests, tired of manual reviews. Indie hackers on tight budgets, leveraging standard or minimal profiles for daily token limits.

Verdict

Worth forking from the openclaw GitHub repo if you're deep in OpenClaw—solid docs and CLI like orchestrate in solo/parallel/full modes deliver real workflow gains, despite 12 GitHub stars signaling early maturity. 0.8% credibility score reflects niche appeal; test on small tasks before production.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.