AI Harness Engineering All In One

AI 马具工程 / AI 驾驭工程

Harness Engineering

Agent = Model + Harness => Harness Engineering = AI Agent - LLM

Harness engineering: leveraging Codex in an agent-first world | OpenAI

AGENTS.md
ARCHITECTURE.md
docs/
├── design-docs/
│   ├── index.md
│   ├── core-beliefs.md
│   └── ...
├── exec-plans/
│   ├── active/
│   ├── completed/
│   └── tech-debt-tracker.md
├── generated/
│   └── db-schema.md
├── product-specs/
│   ├── index.md
│   ├── new-user-onboarding.md
│   └── ...
├── references/
│   ├── design-system-reference-llms.txt
│   ├── nixpacks-llms.txt
│   ├── uv-llms.txt
│   └── ...
├── DESIGN.md
├── FRONTEND.md
├── PLANS.md
├── PRODUCT_SENSE.md
├── QUALITY_SCORE.md
├── RELIABILITY.md
└── SECURITY.md

demos

Agent = Model + Harness

https://www.bilibili.com/video/BV12LR1B3EUt/

Learn Harness Engineering

Harness 的核心机制

核心概念：

• 用明确的规则和边界约束 Agent 的行为。
• 在跨会话的长时任务中保持上下文连续性。
• 防止 Agent 提前宣告任务完成。
• 让 Agent 学会通过完整的流水线测试来验证自己的工作。
• 让 Agent 的运行过程可观测、可调试。

https://walkinglabs.github.io/learn-harness-engineering/en/

https://github.com/walkinglabs/learn-harness-engineering

refs

https://openai.com/index/harness-engineering/

https://martinfowler.com/articles/harness-engineering.html

https://addyosmani.com/blog/agent-harness-engineering/

https://www.anthropic.com/engineering/harness-design-long-running-apps
https://www.anthropic.com/engineering/effective-harnesses-for-long-running-agents

©xgqfrms 2012-2021

www.cnblogs.com/xgqfrms 发布文章使用：只允许注册用户才可以访问！

原创文章，版权所有©️xgqfrms, 禁止转载 🈲️，侵权必究⚠️！