skills.rest
china-qijizhifengchina-qijizhifengOfficial·3 Agent Skills Included

agentic-harness-engineering

Automatically evolve and audit coding agent harnesses with evidence

Evolves coding agent harnesses automatically by improving system prompts, tool descriptions, middleware, skills, and memory while the base model stays fixed. Audits agent workspaces against the HARNESS.md seven-component standard and records every change as a falsifiable, verifiable manifest. Eliminates guesswork tuning by tying each edit to trace evidence, root causes, and predicted impact.
812|90
↗
npx skills add china-qijizhifeng/agentic-harness-engineering --all -g -y

All Skills in This Repository (3)

Pure Emerald Level Indicators
📦 In Repo
china-qijizhifengchina-qijizhifeng

agent-debugger-cli

Configure AI models, download Langfuse traces, and run quality checks.

Official
Intermediate
📦 In Repo
china-qijizhifengchina-qijizhifeng

nexau-evolution-guide

Document NexAU agent evolution with YAML config schemas and component guides.

Official
Advanced
📦 In Repo
china-qijizhifengchina-qijizhifeng

agentic-harness-engineering

Audit and validate coding agent harness structures against HARNESS.md v1.0.

Official
Advanced

Frequently Asked Questions

FAQPage Schema
How to install agentic-harness-engineering?▼

Run `npx skills add china-qijizhifeng/agentic-harness-engineering --all -g -y` in your terminal to install all skills in this suite globally.

What is an agent harness?▼

A harness is everything around the AI model that makes an agent work: system prompts, tool descriptions, tool code, middleware, skills, sub-agents, and long-term memory. This suite standardizes and improves those components.

How does AHE improve coding agents?▼

It runs repeated evaluate-analyze-improve loops where each change is backed by trace evidence and a predicted impact, then verified or rolled back in the next iteration.

Can I audit my existing agent workspace?▼

Yes. The included skill checks your workspace against the HARNESS.md seven-component standard and outputs a compliance score with concrete improvement suggestions.

Which agent frameworks does it support?▼

It ships validation profiles for Hermes, OpenClaw, and Codex, and the HARNESS.md standard works with any agent that uses file-based prompts, tools, and skills.

Related Repositories in Software Engineering

View All in Software Engineering→
openclawopenclaw
388.4k

openclaw

Run a personal AI assistant across your devices and chat apps

Official📦 120 Skills
obraobra
280.0k

superpowers

Gives coding agents a disciplined workflow from idea to merged code

Community📦 15 Skills
facebookfacebook
248.4k

react

AI agent skills for building, testing, and porting React core

Official📦 14 Skills