Agentic Harness Engineering（论文与实现）

AHE (Agentic Harness Engineering) is an open observability system for automatically evolving the harness around a coding agent. The base model is held fixed; what evolves are the harness components — system prompts, tool descriptions, tool implementations, middleware, skills, sub-agents, and long-term memory.

课时

• agents
  • code_agent_simple
    • Long-Term Memory
    • code_agent_simple
    • Short-Term Memory
    • systemprompt
  • evolve_agent
    • Context Compaction Prompt
    • Core Principles
    • middleware
      • context_compaction
        • prompts
          • Emergency Context Compaction Prompt
    • skills
      • agent-debugger-cli
        • _source
          • agent_debugger_core
            • runtime
              • system prompt
        • Agent Debugger CLI
      • nexau-evolution-guide
        • reference
          • 🤖 Core Concepts: Agents
          • hooks
          • 🧠 Core Concepts: LLM Configuration
          • Sandbox System
          • Skills
          • 🛠️ Core Concepts: Tools
        • NexAU Evolution Guide — Simple Agent Starting Point
    • tools
      • nexau 工具实现
  • explore_agent
    • source_agent
      • Context
    • web_agent
      • Context
• experiments
  • evolved_harness
    • Long-Term Memory
    • code_agent_simple
    • Short-Term Memory
    • systemprompt
• Agentic Harness Engineering：以可观测性驱动的编码 Agent Harness 自动演化
• Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses
• skills
  • agentic-harness-engineering
    • references
      • analysis
        • Task Analysis: {{task_id}}
        • Evaluation Overview — Iteration {{iteration}}
      • HARNESS.md Directory Template
      • docs
        • results.json — 评估结果标准格式
        • 回滚策略 — Harness Change Rollback
      • examples
        • trace-sample
          • Evaluation Overview — Iteration 3
          • Trace Sample — Task T-042: Search Results Exceed 50
      • HARNESS.md — Agent Harness Specification v1.0
      • 🦁 HARNESS.md + AHE.skill
    • AHE Skill — Agent Harness Evolution

开始学习 → Long-Term Memory