{"id":50897707,"url":"https://github.com/CrewBench/CrewBench","last_synced_at":"2026-07-03T16:01:19.189Z","repository":{"id":333630612,"uuid":"1138111404","full_name":"CrewBench/CrewBench","owner":"CrewBench","description":"CrewBench is a desktop UI and control layer for AI coding agents such as Claude Code, Gemini CLI, and Codex.","archived":false,"fork":false,"pushed_at":"2026-01-23T07:50:21.000Z","size":298154,"stargazers_count":4,"open_issues_count":2,"forks_count":0,"subscribers_count":2,"default_branch":"main","last_synced_at":"2026-01-23T22:42:16.367Z","etag":null,"topics":["ai","ai-agents","claude-code","claude-code-plugin","cowork","crewbench","gemini-cli","webui"],"latest_commit_sha":null,"homepage":"","language":"TypeScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/CrewBench.png","metadata":{"files":{"readme":"readme.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-01-20T09:00:08.000Z","updated_at":"2026-01-23T07:50:25.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/CrewBench/CrewBench","commit_stats":null,"previous_names":["crewbench/crewbench"],"tags_count":4,"template":false,"template_full_name":null,"purl":"pkg:github/CrewBench/CrewBench","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CrewBench%2FCrewBench","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CrewBench%2FCrewBench/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CrewBench%2FCrewBench/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CrewBench%2FCrewBench/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/CrewBench","download_url":"https://codeload.github.com/CrewBench/CrewBench/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CrewBench%2FCrewBench/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":35092185,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-07-03T02:00:05.635Z","response_time":110,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai","ai-agents","claude-code","claude-code-plugin","cowork","crewbench","gemini-cli","webui"],"created_at":"2026-06-16T01:31:30.075Z","updated_at":"2026-07-03T16:01:19.163Z","avatar_url":"https://github.com/CrewBench.png","language":"TypeScript","funding_links":[],"categories":["ai"],"sub_categories":[],"readme":"\u003ca id=\"readme-top\"\u003e\u003c/a\u003e\n\n[![Contributors][contributors-shield]][contributors-url]\n[![Forks][forks-shield]][forks-url]\n[![Stargazers][stars-shield]][stars-url]\n[![Issues][issues-shield]][issues-url]\n[![project_license][license-shield]][license-url]\n\n\u003cbr /\u003e\n\u003cdiv align=\"center\"\u003e\n  \u003ca href=\"https://github.com/CrewBench/CrewBench\"\u003e\n    \u003cimg src=\"resources/icon.png\" alt=\"Logo\" width=\"80\" height=\"80\"\u003e\n  \u003c/a\u003e\n\n\u003ch3 align=\"center\"\u003eCrewBench\u003c/h3\u003e\n\n  \u003cp align=\"center\"\u003e\n    An open-source, enhanced version of Claude Cowork built specifically for programmers.\n    \u003cbr /\u003e\n    \u003ca href=\"https://github.com/CrewBench/CrewBench\"\u003e\u003cstrong\u003eExplore the docs »\u003c/strong\u003e\u003c/a\u003e\n    \u003cbr /\u003e\n    \u003cbr /\u003e\n    \u003ca href=\"#getting-started\"\u003eGetting Started\u003c/a\u003e\n    \u0026middot;\n    \u003ca href=\"https://github.com/CrewBench/CrewBench/releases\"\u003eView Demo\u003c/a\u003e\n    \u0026middot;\n    \u003ca href=\"https://github.com/CrewBench/CrewBench/issues/new?template=bug_report.yml\"\u003eReport Bug\u003c/a\u003e\n    \u0026middot;\n    \u003ca href=\"https://github.com/CrewBench/CrewBench/issues/new?template=feature_request.yml\"\u003eRequest Feature\u003c/a\u003e\n    \u0026middot;\n    \u003ca href=\"#contact\"\u003eContact Us\u003c/a\u003e\n  \u003c/p\u003e\n\u003c/div\u003e\n\n\u003c!-- TABLE OF CONTENTS --\u003e\n\u003cdetails\u003e\n  \u003csummary\u003eTable of Contents\u003c/summary\u003e\n  \u003col\u003e\n    \u003cli\u003e\n      \u003ca href=\"#overview\"\u003eOverview\u003c/a\u003e\n      \u003cul\u003e\n        \u003cli\u003e\u003ca href=\"#core-features\"\u003eCore Features\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\n          \u003ca href=\"#special-features\"\u003eSpecial Features\u003c/a\u003e\n          \u003cul\u003e\n            \u003cli\u003e\u003ca href=\"#file-level-version-tracking-and-crewbench-logging\"\u003eFile Level Version Tracking\u003c/a\u003e\u003c/li\u003e\n            \u003cli\u003e\u003ca href=\"#action-logs\"\u003eAction Logs\u003c/a\u003e\u003c/li\u003e\n          \u003c/ul\u003e\n        \u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#technical-focus-areas\"\u003eTechnical Focus Areas\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#our-relationship-to-aionui\"\u003eRelationship to AionUi\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#what-crewbench-does-not-do\"\u003eWhat CrewBench Does Not Do\u003c/a\u003e\u003c/li\u003e\n      \u003c/ul\u003e\n    \u003c/li\u003e\n    \u003cli\u003e\n      \u003ca href=\"#getting-started\"\u003eGetting Started\u003c/a\u003e\n      \u003cul\u003e\n        \u003cli\u003e\u003ca href=\"#system-requirements\"\u003eSystem Requirements\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#prerequisites\"\u003ePrerequisites\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#installation\"\u003eInstallation\u003c/a\u003e\u003c/li\u003e\n      \u003c/ul\u003e\n    \u003c/li\u003e\n    \u003cli\u003e\n      \u003ca href=\"#crewbench-usage\"\u003eCrewBench Usage\u003c/a\u003e\n      \u003cul\u003e\n        \u003cli\u003e\u003ca href=\"#cowork-with-cli-ai-agents\"\u003eCowork With CLI AI Agents\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#cowork-directly-on-your-codebase\"\u003eCowork Directly on Your Codebase\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#inspect-agent-output\"\u003eInspect Agent Output\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#file-timeline-and-action-logs\"\u003eFile Timeline and Action Logs\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#parallel-task-execution\"\u003eParallel Task Execution\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#remote-coworking\"\u003eRemote Coworking\u003c/a\u003e\u003c/li\u003e\n        \u003cli\u003e\u003ca href=\"#developer-controlled-interface\"\u003eDeveloper-Controlled Interface\u003c/a\u003e\u003c/li\u003e\n      \u003c/ul\u003e\n    \u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#roadmap\"\u003eRoadmap\u003c/a\u003e\u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#contributing-to-crewbench\"\u003eContributing\u003c/a\u003e\u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#license\"\u003eLicense\u003c/a\u003e\u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#contact\"\u003eContact\u003c/a\u003e\u003c/li\u003e\n    \u003cli\u003e\u003ca href=\"#acknowledgments\"\u003eAcknowledgments\u003c/a\u003e\u003c/li\u003e\n  \u003c/ol\u003e\n\u003c/details\u003e\n\n\u003c!-- ABOUT THE PROJECT --\u003e\n\n## Overview\n\n![CrewBench Home](resources/crewbenchhome.png)\n\nCrewBench is a **desktop UI and control layer for AI coding agents** such as Claude Code, Gemini CLI, and Codex.\n\nIt provides **visibility, control, and rollback** over how AI agents read and modify a real codebase.\n\n\u003e [!NOTE]\n\u003e CrewBench is built on top of **[AionUi](https://github.com/iOfficeAI/AionUi) (Apache-2.0)** and inherits all upstream AionUi capabilities, while adding **developer-focused observability and guardrails**.\n\n## Core Features\n\n| Feature                          | Description                                                                                                        |\n| -------------------------------- | ------------------------------------------------------------------------------------------------------------------ |\n| **Universal Model Support**      | Work with cloud and local models through a unified interface, including OpenAI-compatible APIs and local runtimes. |\n| **Agent Coworking Environment**  | Multiple agents and tasks can run in parallel with isolated context, enabling real cowork-style workflows.         |\n| **Direct File System Access**    | Agents operate directly on real project files via a file-tree interface, not abstract chat outputs.                |\n| **File Timeline**                | Every file change is recorded in a chronological timeline, making agent activity inspectable and reviewable.       |\n| **Action Logs \u0026 Audit Trail**    | Full logs of both AI and human actions provide observability, debugging capability, and reproducibility.           |\n| **Live Preview \u0026 Editing**       | Generated artifacts are surfaced as editable, previewable files rather than hidden chat responses.                 |\n| **Extensible \u0026 Customizable UI** | The interface is developer-controlled, supporting layout flexibility and custom CSS theming.                       |\n\n### Special Features:\n\n#### File Level Version Tracking and CrewBench Logging\n\n![File Timeline](resources/timeline.png)\n\nCrewBench adds **AI-aware file version tracking**, recording every agent-initiated change:\n\n- Full content snapshots per change\n- Line-based diffs between versions\n- Operation type (`create`, `write`, `delete`)\n- Agent attribution\n- Timestamped history per workspace\n\nThis complements Git by tracking **agent behavior**, not commits.\n\n#### Action Logs\n\n![Action Logs](resources/logs.png)\n\nAction logs provide a comprehensive audit trail of all actions performed by agents and humans in the workspace. Each log entry includes:\n\n- Timestamp\n- Action type (create, write, delete)\n- Agent attribution\n- File path\n- Content snapshots\n- Line-based diffs\n\n---\n\n### Technical Focus Areas\n\n| Focus Area                  | What It Covers                                                                                                  | Why It Matters                                                                           |\n| --------------------------- | --------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------- |\n| **Observability**           | • File timelines and diffs\u003cbr\u003e• Clear agent vs human attribution\u003cbr\u003e• Action-level history                      | Lets developers _see exactly what happened_, who did it (human or AI), and in what order |\n| **Control / Guardrails**    | • Path-based read/write restrictions\u003cbr\u003e• Operation-level permissions\u003cbr\u003e• Confirmation gates for risky actions | Prevents agents from touching the wrong files or making unsafe changes                   |\n| **Reproducibility**         | • Inspectable agent action sequences\u003cbr\u003e• Debugging and diagnosis of AI behavior                                | Makes AI behavior debuggable, auditable, and repeatable                                  |\n| **Human + Agent Workflows** | • Continuous history across human and agent edits\u003cbr\u003e• No context loss between steps                            | Enables seamless collaboration without losing state or intent                            |\n\n### Our Relationship to AionUi\n\nAionUi provides the base platform, including:\n\n- Multi-agent orchestration\n- Desktop UI infrastructure\n- File system access\n- Local-first execution\n- Cross-platform support\n\nCrewBench builds **on top of this foundation** and focuses specifically on **software engineering workflows**.\n\n\u003e [!NOTE]\n\u003e\n\u003e ### What CrewBench Does Not Do\n\u003e\n\u003e - Does not replace Git\n\u003e - Does not replace an IDE\n\u003e - Does not introduce new AI models\n\u003e - Does not focus on prompt presets\n\u003e\n\u003e CrewBench operates at the **control and observability layer**, not the editor or model layer.\n\n## Getting Started\n\n\u003cp\u003e\n  \u003ca href=\"https://github.com/CrewBench/CrewBench/releases\"\u003e\n    \u003cimg src=\"https://img.shields.io/badge/Download-Latest%20Release-32CD32?style=for-the-badge\u0026logo=github\u0026logoColor=white\" alt=\"Download Latest Release\" height=\"25\"\u003e\n  \u003c/a\u003e\n\u003c/p\u003e\n\n### System Requirements\n\n- **macOS**: 10.15 or higher\n- **Windows**: Windows 10 or higher\n- **Linux**: Ubuntu 18.04+ / Debian 10+ / Fedora 32+\n- **Memory**: Recommended 4GB or more\n- **Storage**: At least 500MB available space\n\n### Prerequisites\n\n- CrewBench is a desktop application and does not require any prerequisites.\n\n### Installation\n\n1. Download the latest release from [here](https://github.com/CrewBench/CrewBench/releases).\n2. Install the downloaded file.\n3. Now you can run the `CrewBench as an Application` file.\n\n\u003cp align=\"right\"\u003e(\u003ca href=\"#readme-top\"\u003eback to top\u003c/a\u003e)\u003c/p\u003e\n\n## CrewBench Usage\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong id=\"cowork-with-cli-ai-agents\"\u003eCowork With CLI AI Agents (Unified Developer Interface)\u003c/strong\u003e\u003c/summary\u003e\n\nCrewBench acts as a graphical cowork layer on top of existing command-line AI coding tools like Gemini CLI, Claude Code, Codex, Qwen Code, Goose, and more.\n\n| Capability           | Description                                      |\n| -------------------- | ------------------------------------------------ |\n| Multi-Agent Sessions | Run multiple AI coding agents simultaneously     |\n| Unified Interface    | One UI instead of multiple terminals             |\n| Context Isolation    | Each agent and task runs with independent memory |\n| Local Execution      | Agents run locally on your machine               |\n\n\u003c/details\u003e\n\n---\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong id=\"cowork-directly-on-your-codebase\"\u003eCowork Directly on Your Codebase\u003c/strong\u003e\u003c/summary\u003e\n\nAgents operate on real project files — not abstract chat outputs.\n\n| Developer Need    | How CrewBench Supports It                  |\n| ----------------- | ------------------------------------------ |\n| Code Exploration  | File-tree based navigation                 |\n| File Modification | Agents read and write actual project files |\n| Batch Operations  | Rename, merge, and reorganize files        |\n| Safe Changes      | All changes are tracked and reviewable     |\n\n\u003c/details\u003e\n\n---\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong id=\"inspect-agent-output\"\u003eInspect Agent Output, Not Just Results\u003c/strong\u003e\u003c/summary\u003e\n\nCrewBench surfaces agent outputs as first-class artifacts you can inspect and modify.\n\n| Visibility Feature   | Description                                     |\n| -------------------- | ----------------------------------------------- |\n| Multi-Format Preview | Code, Markdown, HTML, documents, images, diffs  |\n| Live File Sync       | Previews update automatically when files change |\n| Inline Editing       | Edit or refine agent output directly            |\n| Change Awareness     | See what changed, not just the final state      |\n\n\u003c/details\u003e\n\n---\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong id=\"file-timeline-and-action-logs\"\u003eFile Timeline and Action Logs\u003c/strong\u003e\u003c/summary\u003e\n\nEngineering-grade observability for AI coding agents.\n\n| Observability Tool | Purpose                                        |\n| ------------------ | ---------------------------------------------- |\n| File Timeline      | Chronological history of every file change     |\n| Action Logs        | Full audit trail of agent and human actions    |\n| Change Attribution | Clear mapping of who (or which agent) did what |\n| Reproducibility    | Inspect and debug agent behavior               |\n\n\u003c/details\u003e\n\n---\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong id=\"parallel-task-execution\"\u003eParallel Task Execution Without Context Collision\u003c/strong\u003e\u003c/summary\u003e\n\nRun multiple tasks simultaneously without cross-contamination.\n\n| Workflow Pattern  | Support                                |\n| ----------------- | -------------------------------------- |\n| Parallel Sessions | Multiple active tasks at the same time |\n| Isolated Memory   | No cross-task confusion                |\n| Long-Running Work | Persistent sessions for ongoing work   |\n\n\u003c/details\u003e\n\n---\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong id=\"remote-coworking\"\u003eRemote Coworking and Team Workflows\u003c/strong\u003e\u003c/summary\u003e\n\nControl and observe agents remotely while keeping execution local.\n\n| Capability          | Description                                           |\n| ------------------- | ----------------------------------------------------- |\n| WebUI Access        | Control CrewBench from other devices on the network   |\n| Local Agent Control | Agents continue running on the host machine           |\n| Server Deployment   | Suitable for self-hosted setups                       |\n| CrewBench Teams     | Collaborative multi-developer workflows (coming soon) |\n\n\u003c/details\u003e\n\n---\n\n\u003cdetails\u003e\n\u003csummary\u003e\u003cstrong id=\"developer-controlled-interface\"\u003eDeveloper-Controlled Interface\u003c/strong\u003e\u003c/summary\u003e\n\nCrewBench adapts to your workflow — not the other way around.\n\n| Customization      | Description                              |\n| ------------------ | ---------------------------------------- |\n| CSS Customization  | Full control over UI appearance          |\n| Layout Flexibility | Shape the interface around your workflow |\n| Personal Setup     | Make CrewBench feel like your own IDE    |\n\n\u003c/details\u003e\n\n\u003cp align=\"right\"\u003e(\u003ca href=\"#readme-top\"\u003eback to top\u003c/a\u003e)\u003c/p\u003e\n\n## Roadmap\n\n- [Completed] File Timeline\n- [In Progress] Action Logs\n- [Not Started] GuardRails\n\n\u003cp align=\"right\"\u003e(\u003ca href=\"#readme-top\"\u003eback to top\u003c/a\u003e)\u003c/p\u003e\n\n\u003c!-- CONTRIBUTING --\u003e\n\n## Contributing to CrewBench\n\nTo contribute to CrewBench, please read the [CONTRIBUTING.md](CONTRIBUTING.md) file.\n\n### Top contributors:\n\n\u003ca href=\"https://github.com/CrewBench/CrewBench/graphs/contributors\"\u003e\n  \u003cimg src=\"https://contrib.rocks/image?repo=CrewBench/CrewBench\" alt=\"contrib.rocks image\" /\u003e\n\u003c/a\u003e\n\n\u003c!-- LICENSE --\u003e\n\n## License\n\nDistributed under the project_license. See `LICENSE.txt` for more information.\n\n\u003cp align=\"right\"\u003e(\u003ca href=\"#readme-top\"\u003eback to top\u003c/a\u003e)\u003c/p\u003e\n\n\u003c!-- CONTACT --\u003e\n\n## Contact\n\nAbhinav Sharma - [x/abhinavsharma_x](https://twitter.com/abhinavsharma_x)\n\nProject Link: [https://github.com/CrewBench/CrewBench](https://github.com/CrewBench/CrewBench)\n\n\u003cp align=\"right\"\u003e(\u003ca href=\"#readme-top\"\u003eback to top\u003c/a\u003e)\u003c/p\u003e\n\n\u003c!-- ACKNOWLEDGMENTS --\u003e\n\n## Acknowledgments\n\n- [Electron](https://www.electronjs.org/) — for enabling cross-platform desktop application development\n- [Node.js](https://nodejs.org/) — for the underlying runtime and tooling ecosystem\n- [SQLite](https://www.sqlite.org/) — for reliable, local-first data storage\n- [OpenAI](https://openai.com/) — for OpenAI-compatible APIs and ecosystem standards\n- [Google Gemini](https://deepmind.google/technologies/gemini/) — for multimodal and large-context model support\n- [Anthropic Claude](https://www.anthropic.com/) — for advancing safe and capable coding agents\n\n\u003cp align=\"right\"\u003e(\u003ca href=\"#readme-top\"\u003eback to top\u003c/a\u003e)\u003c/p\u003e\n\n[contributors-shield]: https://img.shields.io/github/contributors/CrewBench/CrewBench.svg?style=for-the-badge\n[contributors-url]: https://github.com/CrewBench/CrewBench/graphs/contributors\n[forks-shield]: https://img.shields.io/github/forks/CrewBench/CrewBench.svg?style=for-the-badge\n[forks-url]: https://github.com/CrewBench/CrewBench/network/members\n[stars-shield]: https://img.shields.io/github/stars/CrewBench/CrewBench.svg?style=for-the-badge\n[stars-url]: https://github.com/CrewBench/CrewBench/stargazers\n[issues-shield]: https://img.shields.io/github/issues/CrewBench/CrewBench.svg?style=for-the-badge\n[issues-url]: https://github.com/CrewBench/CrewBench/issues\n[license-shield]: https://img.shields.io/github/license/CrewBench/CrewBench.svg?style=for-the-badge\n[license-url]: https://github.com/CrewBench/CrewBench/blob/master/LICENSE.txt\n[product-screenshot]: images/screenshot.png\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FCrewBench%2FCrewBench","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FCrewBench%2FCrewBench","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FCrewBench%2FCrewBench/lists"}