mini-harness

Better Harness is an open-source Harness Engineering platform for coding agents. It analyzes project and, where supported, session evidence around an agent's work rather than only its final code diff, evaluating the Agent Work Loop for issues such as task understanding, validation, delivery control, and retained lessons. It turns supported gaps into prioritized findings linked to their evidence, expected outcomes, repair boundaries, and acceptance checks, and produces host-specific reports in formats including HTML, paired Markdown, or native Canvas reports. The platform also supports defining harnesses as code, running controlled experiments, inspecting evidence, and comparing outcomes across coding-agent hosts such as Claude Code, Codex, Qoder, Cursor, and GitHub Copilot CLI.

uni-robot-agent is an open-source harness for robot-use agents, providing shared robot-tool interfaces across simulated environments and runtimes including Codex and Claude Code. It can run a built-in agent, invoke robot tools from a custom Python strategy, and inspect live or saved episodes. Its executor schedules agent and SDK tool calls, while adapters connect the shared tools to native environment backends; Direct mode uses robot tools, and Hybrid mode can add a compatible action model such as Pi0.5. The package includes a Python SDK, CLI, prompts, skills, and a dashboard that displays camera streams, tool calls, execution timing, and task outcomes, with replay controls for saved runs. It currently supports simulation integrations including RoboDojo, RoboUse, and RoboLab; supported physical-robot deployment is listed as forthcoming. The package is installable with pip or uv and is licensed under the MIT License.

Harness² is an open-source framework for recursively optimizing the system around an AI model while keeping the model weights fixed. At the task level, it uses a contrastive Propose–Probe–Compose process: it proposes edits to an execution harness, probes their effects on the same task, and composes a refined harness from behavioral differences between executions without ground-truth feedback. A second recursion level maintains an improvement harness containing adopted harnesses and evidence-supported lessons for future tasks. Its editable components include system prompts, guardrails, memory, tool guidance, subagents, skills, scripts, and plugins. The repository supports parallel and sequential recursion, OpenCode and Codex executors, and benchmarks including JobBench, LAB, and WorkBuddy-Bench. It runs locally with Linux, Python 3.12, and uv, and is licensed under Apache 2.0; the repository states that it is not an officially supported Google product.