openharness
zhijiewong/openharness
OpenHarness is an extensible, provider-agnostic terminal AI coding agent that runs locally with Ollama or via cloud APIs, featuring a rich TUI and 44 built-in tools.
96 23MITUpdated 2026-05-12
72
GoodOverview
OpenHarness is a powerful terminal-based AI coding assistant designed to give developers freedom of choice regarding LLM providers, supporting both free local models (via Ollama) and major cloud APIs. It offers a comprehensive set of 44 built-in tools, permission controls, session management, checkpoints with rewind functionality, and a feature-rich terminal UI complete with syntax highlighting, auto-completion, and customizable themes.
Capabilities
- ▸Provider-agnostic LLM support (Ollama and cloud APIs)
- ▸44 built-in tools covering file manipulation, web search, task management, and sub-agents
- ▸Interactive terminal user interface with markdown rendering and syntax highlighting
- ▸Git-backed checkpointing and file change rollback (/rewind)
- ▸Headless mode and CI/CD integration with JSON output
- ▸Python and TypeScript SDKs for programmatic agent control
Best for
Developing and debugging code locally using free offline LLMs, Running automated code review and testing tasks in headless CI/CD environments, Executing multi-file edits, refactoring, and codebase searches directly from the terminal, Managing complex multi-agent coding workflows and task pipelines
Works with
Claude CodeCursorGitHub Copilot
Evoa Score breakdown= Σ (score × weight)
72Task usefulness20%100 → +20.0
実タスクにどれだけ役立つか（機能の豊富さ・用途の明確さ）。 — AIによるcapabilities/use-cases解析
Code quality15%90 → +13.5
実装・指示の品質。 — AIによるSKILL.md/README解析
Maintenance15%70 → +10.5
リポジトリがどれだけ活発に保守されているか。 — GitHub 最終push日時の新しさ
Documentation12%34 → +4.1
ドキュメントの充実度・分かりやすさ。 — README/独自要約の情報量
Security15%50 → +7.5
危険・不審な挙動が無いか。 — AIによるセキュリティレビュー
Originality10%85 → +8.5
ありふれたラッパーではない独自性。 — AIによる独自性判定
Popularity8%54 → +4.3
コミュニティの採用度。 — GitHub Stars/Forks（対数スケール）
Compatibility5%75 → +3.8
対応AIエージェントの広さ。 — AIによる対応エージェント判定
Weighted total72.2 / 100
ライセンス不明/制限あり（Red）のSkillは総合スコアに0.85倍の補正を適用します。 ランキングはこのScoreのみで決まり、広告で変わりません。 算出方法の詳細 →
Security considerations
OpenHarness executes system-level tools such as bash commands, remote triggers, and process killers. High-risk actions require user approval by default or can be automated using trust/auto modes.
Categories
Summary and analysis are original content generated by AI Skills Rank. The skill's source text is not reproduced here — view it on the linked repository.