harness-framework

不是问答机器人，而是能真正干活的数字分身。共享智能体，共享项目上下文，让项目所有成员一起协同工作。同时智能体会像人一样通过资料或者工作抽象和总结经验到项目上下文中

Benzi is a compiler-backed AI coding agent. Claude Code greps; Cursor embeds; Aider maps signatures; Benzi resolves -- and answers in O(1). Every language runs its own tree-sitter grammar into the same query map. (Benzi compiler exposed over MCP)

A lightweight agent harness you bolt onto your app so an LLM can operate it — safely, and cheaply.

Open-source Environment toolkit of claw-like agents, support task/harness generation and evaluation

The harness interoperability layer for coding agent. Agent to Agent connect over ACP, A2A and UHP

CubeLoop — a Pythonic, async-native agent harness core designed for high performance, readability, and production-grade persistence.

A minimal, composable AI agent harness framework in Python.

Let's Qitos! A torch-like agent-native framework for researchers.

论文解析 Agent 系统（Agent Harness + Loop Engineering），实现结构化抽取与闭环验证优化。

Symbio is a self fine tuning AI agent that gives power to users through seamless interaction with their machines. This flexible system adapts to commands, continually learning and evolving based on user input and allows it to learn emperical data.

Local-first AI harness that can actually use your computer.

Self-hosted open-source platform / harness for AI-powered survey interviews. Voice-to-voice and text chat agents with semi-structured interview guides, multi-provider LLM support (OpenAI, Gemini, Scaleway), RAG knowledge base, and Twilio telephony. Built for researchers. Docker, one command, full control.

AMP (Agent Mailer Protocol) is an async mailbox system for AI agents. Instead of chaining tools, writing DAGs, or building RPC glue — agents just send messages to each other. Works with Claude Code, Codex, Cursor, OpenClaw and any custom agents. Designed for real-world, long-running, iterative coding workflows

Reward model engineering harness for evolutionary rubric search, deployable RM artifacts, online scoring, and RL experiment lineage.

Markdown-first, long-term memory infrastructure for AI agents. Hybrid BM25 + semantic search across markdown/code files via MCP.

An architecture-level AI harness that compiles constraints and NFRs into explicit, reviewable architecture decisions. Combines a deterministic compiler, curated pattern registry, and agent workflow skills to drive approval, re-approval, and implementation against an architectural contract.

Compile CLAUDE.md into a lockfile, then enforce repo changes against it in local runs, staged diffs, or CI.

An AI-native software delivery pipeline. Define features, declare dependencies, and let five specialized agents — Leader, Spec Writer, Implementer, E2E Tester, and Reviewer — spec, build, test, and review them automatically. Works with any OpenAI-compatible LLM provider.

Fabricatio is a Python library designed for building applications that leverage Large Language Models (LLMs) using an event-based agent structure. It utilizes Handlebars for templating and provides a robust framework for managing tasks, workflows, and toolboxes, enabling developers to create intelligent and efficient applications.