HKUDS has released OpenHarness, an open-source agent harness framework in Python that includes a built-in personal agent called Ohmo. The project has accumulated nearly 15,000 GitHub stars, suggesting meaningful community traction. It appears to be a general-purpose scaffolding layer for building and evaluating AI agents.
oh-my-openagent (omo) is a TypeScript-based agent harness project on GitHub, previously known as oh-my-opencode. The repository has accumulated 58,729 stars with 180 new stars today, indicating significant community traction. The rename from 'opencode' to 'openagent' suggests a broadening scope beyond code-focused tasks toward general agent capabilities.
A GitHub repository aggregating resources on AI agent harness engineering, covering tools, patterns, evaluations, memory systems, MCP (Model Context Protocol), permissions, observability, and orchestration. The list has accumulated 1,318 stars with 39 added today, indicating moderate community traction. It serves as a reference index rather than original research or tooling.
This survey paper introduces the concept of 'code as agent harness,' framing code not merely as output but as the operational infrastructure for LLM-based agents—covering reasoning, action, environment modeling, and execution-based verification. The authors organize the analysis across three layers: harness interface, harness mechanisms (planning, memory, tool use, feedback control), and scaling to multi-agent systems. Applications span coding assistants, GUI/OS automation, embodied agents, scientific discovery, and enterprise workflows. Open challenges include evaluation beyond task success, verification under incomplete feedback, and human oversight for safety-critical actions.
NanoBot is a lightweight, open-source Python AI agent framework from HKUDS (Hong Kong University of Science and Technology) designed for tool use, chat, and workflow automation. The repository has accumulated 45,658 stars with 120 added today, indicating significant community traction. It competes in the growing space of open-source agent harnesses.
Andrew Ng and collaborators Rohit Prasad and Devika Verma have released OpenCoworker, a free open-source desktop agent built by extending the aisuite library to support agent harnesses. The tool allows users to connect frontier LLMs (OpenAI, Anthropic, Google) or local models via Ollama to desktop tasks including file access, messaging, and workflow automation, with privacy as a design priority. Ng frames this as a response to data-retention concerns with commercial desktop agents, citing Anthropic's Fable release as a recent example of policy opacity. The post also provides a concise overview of the current desktop agent landscape and the shift toward LLM-driven agentic loops.
A GitHub repository called 'agents' by wshobson provides a multi-harness agentic plugin marketplace targeting Claude Code, Codex CLI, Cursor, OpenCode, GitHub Copilot, and Gemini CLI. The project has accumulated 37,134 stars with modest daily momentum (+43 today). It represents a cross-platform approach to agent tooling that spans multiple competing AI coding environments.
OpenAI has released an updated version of its Agents SDK with TypeScript support, adding sandbox agent capabilities and a built-in open-source harness. The update expands the SDK's language coverage beyond Python and lowers the barrier for TypeScript-based agent development. This is a steady-state ecosystem tooling update from a tier-1 source.
OpenHands, an open-source AI-driven software development platform implemented in Python, is trending on GitHub with 77,048 total stars and 258 new stars today. The project enables AI agents to perform software development tasks autonomously. Its continued traction signals sustained community interest in open-source coding agent frameworks.