🐬DeepChat - A smart assistant that connects powerful AI to your personal world
An open-source Harness Engineering platform for coding agents—define harnesses as code, run controlled experiments, inspect evidence, and compare outcomes.
Approving is an open-source, self-hostable platform for turning coding agents into visual, reviewable, and recoverable delivery workflows. Agents run in real Docker sandboxes, exchange structured artifacts, and pause for human Approve at critical nodes.
A self-improvement loop for coding agents. Mines failure traces, proposes harness edits, validates against a held-out split, and never merges its own work.
Claude Code skill that diagnoses and improves harness configurations based on 8 design principles from Anthropic's harness design article
Drop in any model. Get a harness. Offline-first CLI + TUI that detects model type, runs the right harness, serves HTTP, and scaffolds plugins — for tabular, embeddings, LLMs, vision, diffusion, computer use, and more.
This is a best practice for managing the DeepSeek Harness project collection. You can directly fork the entire repository, rename the `project-template` folder to your own workspace name, open the workspace, and start having fun with DeepSeek Harness.
Agent-ready, evolvable full-stack project generator for reliable AI coding and vibe coding. Scaffold governed FastAPI + Vue 3 repositories with Codex skills, architecture rules, validation harnesses, Docker, CI, auth, i18n, PostgreSQL, and Redis Streams.
Turn a single sentence into a full-stack web app. OPPE orchestrates AI agents (Planner → Builder → QA) to plan, generate, and verify production-ready code from your idea.
This is the corresponding code for the O'Reilly book: Harness Engineering
Smallest readable coding-agent harness that scores on benchmarks. ~970 lines, 59.6% on Terminal-Bench 2.0.
J.A.R.V.I.S official custom harness marketplace - enabling Jarvis to control more desktop software.
Caelestis Architecture Model
Agent Harness 是一个面向企业级 AI Agent 开发治理的本地化框架与模板仓库，提供从需求分析、任务编排、工具权限、MCP 配置、审批流程、轨迹评估、成本记录到回归验证的完整治理骨架。当前项目已完成本地 MVP 与 Hardening 骨架，支持模拟试点验证、质量门禁和变更归档；真实外部试点、MCP 进程启动和远端 CI 可在获得项目与权限后继续接入。
Public-safe agent evaluation write-ups: harness gates, multi-path coding screens, task-family transfer, RAG routes, agent-as-evaluator. Guided Pages tour.
Scaffolding for designing human-AI collaboration architecture. Companion to aiharness.design.
GAN-inspired adversarial development loop — Planner directs Generator + Evaluator in fresh cmux panes, gated by hard pass/fail per deliverable. Adapted from Anthropic's harness-design guidance for long-running agentic apps.
An issue-centric framework for automated agent design or harness evolution
Practical connector selection for rugged, industrial, military-style, and electromechanical systems.
Add a description, image, and links to the harness-design topic page so that developers can more easily learn about it.
To associate your repository with the harness-design topic, visit your repo's landing page and select "manage topics."