# Repo-to-Agent-Context 面向 AI 编程 Agent 的 Context Harness：将代码仓库转化为可读、可检索、可更新、可验证的任务级上下文资产，帮助 Codex / Claude Code / Cursor 减少 token 浪费、上下文丢失、误改与回归风险。 它不是另一个代码生成 Agent，而是 Agent Harness Engineering 基础设施：围绕“任务计划、上下文打包、编辑边界、影响分析、测试推荐、验证报告、上下文更新”构建 Loop Engineering 闭环，让编码 Agent 在复杂仓库中更少瞎猜、更少乱改、改完知道如何验证。

Harness AI & Agent OS for governed multi-agent systems in Python — Nexus orchestration, policy, tools, skills, RAG & multimodal evaluation. Prototype fast, harden agents for production.

Pluggable DeepEval scaffold for RAG, agents, and LLM apps across Anthropic, Bedrock, Azure OpenAI, and Vertex. Ships traceability, test synthesis, safety/PII gating, multi-turn conversation eval, agentic tool-use scoring, JSON validation, judge benchmarks, hyperparameter sweeps, and pytest CI — one Makefile target per feature.

Drop-in TruLens evaluation harness for tool-calling LangGraph agents. Swap LLM providers (OpenAI, Anthropic via LiteLLM, Bedrock, Cortex, Gemini, Ollama) with a single env var. Ships with the RAG Triad plus Plan Quality, Plan Adherence, Execution Efficiency, and Logical Consistency metrics.

Provider-agnostic RAG evaluation harness powered by RAGAS with pluggable LLM and embedding backends.

enjoy harness ai, auto drive, Autonomous harness driving

Building AI developer tools in Rust 🦀 | MCP servers • Agentic CI/CD • Code Intelligence • Hodei Platform

O Agent Harness é um framework de orquestração multi-agente com memória em camadas, busca híbrida (FTS5 + vetorial), e controles de segurança com human-in-the-loop. O sistema implementa uma arquitetura Agente Pai (Orquestrador) + Sub-Agentes Especialistas, onde o Orquestrador nunca executa tarefas diretamente — apenas delega para especialistas