unity-cli
Skill178 starsChanged 32 days agoMITView on GitHub
使用本项目 Unity CLI 查询和验证 Editor/Development Player,操作场景、Prefab、组件和资源数据库,执行编译、测试和构建预检。涉及实际 Unity 状态、CLI 安装或连接排查时使用;纯源码或文档任务不连接 Editor。
6 other projects use a version of this.
Adopted by
Most recently taken up by
- badjano/skillstheir own version, first seen Oct 5, 2026
- Unity-Technologies/skillstheir own version, first seen Sep 22, 2026
- niqibiao/unity-cli-skilltheir own version, first seen Sep 20, 2026
- keijiro/unity-skillstheir own version, first seen Sep 16, 2026
- DennysOliveira/unity-clitheir own version, first seen Aug 31, 2026
- ZHAO0424/unity-cli-skilltheir own version, first seen Aug 5, 2026
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 6
- Unity-Technologies/skills★ 1.2ksame name
- niqibiao/unity-cli-skill★ 43same name
- keijiro/unity-skills★ 11same name
- ZHAO0424/unity-cli-skill★ 2same name
- badjano/skills★ 1same name
- DennysOliveira/unity-cli★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository7 other files it gives its agents.
- AGENTS.mdAGENTS.md
- GameUnity/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- GameUnity/CLAUDE.mdCLAUDE.md
- dgame-devSkill
- luban-devSkill
- skill-creatorSkill
The file
Files
DGame Unity CLI
执行入口
Unity CLI 原生命令通过客户端已加载的 MCP 工具或 unity command <name> 调用。MCP 工具已暴露时直接使用其描述与参数,先以 editor_status.projectPath 核对当前工程;不要求服务名为 unity-dgame,也不要求每次读取客户端配置或运行本地 CLI。任务需要 CLI 且项目副本缺失时,先运行 python .agents/scripts/workflow.py doctor 完成准备和验收,再继续调用。未暴露对应工具时,直接 CLI 固定 --project-path <GameUnity 绝对路径>;Player 使用明确的 --runtime-path。
直接 CLI 使用 GameUnity/Tools/unity.exe,缺失时按 版本配置 准备项目副本。Pipeline 0.10 使用分组命令,例如 unity command --project-path <绝对路径> editor status;完整能力目录用 pipeline list_commands --detail full。CLI 不是 Unity Editor 的同名程序。
MCP 工具由客户端配置的 unity mcp 服务提供,Pipeline 为原生命令提供描述和参数 Schema;技能不注册工具,Workflow 也不负责 MCP 发现。agents/openai.yaml 的自动调用策略用于技能选择,不设置工具优先级。普通原生命令不需要 Workflow 白名单、包装脚本或审批文件。
查询、验证、安装与 MCP 接入统一见 操作规程,按任务读取对应章节。组合验证、CI 和报告按需使用 项目 Workflow,Luban 生产导表使用 luban-dev。
| 操作 | CLI 命令 | MCP 工具 |
| 工程与场景查询 | editor status、scene list_open | editor_status、scene_list_open |
| 脚本编译 | test status → script recompile → script recompile_status | test_status → script_recompile → script_recompile_status |
| 测试 | test list → test status → test run → test status | test_list → test_status → test_run → test_status |
不变量
- 只绑定当前 DGame 工程;导入、编译、测试与写入串行。新操作前核对唯一目标及状态,查询和轮询允许观察忙操作;直接 MCP/CLI 不经过 Workflow 工程锁。
- 源码用代码编辑工具修改,对象、序列化与资产用已注册命令;不通过文本写入、任意 eval、菜单或 Scene/Prefab YAML 绕过编辑和授权规则。
- 普通操作按用户任务及已有授权执行;先明确目标和影响范围,支持
dry_run时直接预览,不支持时读取目标并说明变更。删除、覆盖、依赖、设置、平台和真实构建按仓库授权边界处理,confirm不代替用户授权。 - 按命令真实实现核对副作用、
dry_run/confirm与 Undo 能力,执行后回读目标。超时先检查状态,禁止换入口自动重放写入。 - 编译或测试启动前查询
test_status;已有测试运行时等待,不依赖editor_status=ready判断测试空闲,不取消或覆盖已有测试。 - 异步命令检查触发响应,在调用方设定的总体期限内轮询终态;编译/测试默认每阶段 300 秒,超期报告 blocked/结果未知并保留现场,不自动取消或重放。测试先发现用例,按任务选择模式和筛选器,核对执行数与结果。零用例、全跳过、失败、取消或旧结果不能通过。
- 连接、命令或必需依赖缺失时报告具体 blocked 原因;任务需要项目 CLI 且副本缺失时先按安装验收流程准备,准备失败后再 blocked。不自动升级全局 CLI、启动 Editor、修改用户配置或借用其他工程。Pipeline 鉴权信息不得输出。
权威资料
按任务读取 GameUnity/Packages/com.unity.pipeline/Documentation~/:
| 场景 | 包内文档 |
| 连接、鉴权、Player | connectivity.md、runtime-setup.md |
| 路径、预览、Undo | safety-and-mutations.md、authoring-commands.md |
| 编译、测试、构建 | commands/build-and-compilation.md |
| 资源、对象、Prefab | commands/assets-and-files.md、commands/gameobjects-and-components.md、commands/prefabs.md |
| 截图、场景、设置 | commands/capture.md、commands/scenes.md、commands/project-settings.md |
| 新增项目命令 | creating-commands.md |
实时工具描述、CLI 帮助和源码优先于示例。截图需核对真实视图及输出路径,不等于视觉验收;业务框架 API 使用 dgame-dev。
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

