app-developer
Skill930 starsChanged 6 days agoMITView on GitHub
Build SoduWork apps and skills.
In this repository38 other files it gives its agents.
- AGENTS.mdAGENTS.md
- agent-product-designSkill
- brainstormingSkill
- browser-useSkill
- codebase-designSkill
- code-reviewSkill
- computer-useSkill
- delivery-checkSkill
- diagnosing-bugsSkill
- dispatching-parallel-agentsSkill
- domain-modelingSkill
- executing-plansSkill
- finishing-a-development-branchSkill
- frontend-designSkill
- grillingSkill
- grill-with-docsSkill
- improve-codebase-architectureSkill
- mcp-studioSkill
- mcp-toolsSkill
- officeSkill
- pptxSkill
- product-incubatorSkill
- prototypeSkill
- receiving-code-reviewSkill
- requesting-code-reviewSkill
- resolving-merge-conflictsSkill
- skill-studioSkill
- subagent-driven-developmentSkill
- systematic-debuggingSkill
- taskSkill
- tddSkill
- test-driven-developmentSkill
- using-git-worktreesSkill
- using-superpowersSkill
- verification-before-completionSkill
- wizardSkill
- writing-plansSkill
- writing-skillsSkill
The file
Files
即用开发
目标:精致简洁、打开即用。英文 App,ID稳定。
边界:宿主提供通用能力;群组、成员关系、归组等业务由具体即用实现。
界面:中英文、一个契合业务的主题,不强制明暗双模式。
产品:意图→AI→成果;只收必要输入,AI起草/归类/标签,复制/搜索/游戏直接执行。不以大量表单+“AI帮写”冒充AI产品。
执行 development-workflow.md:设计→实现→检查→安装打开→写回→重开/升版;缺证据不报通过。
- 设计:design-quality.md;GitHub调研/许可;视觉调研可选,查最新同类。
- 单个彩色图标:icon-design.md;配色自由、方形饱满、小尺寸清楚;顶层 icon=PNG≤256px/96KB;查明暗背景,不强制整套/灰色版。
- 五类:instant-types.md;持续角色:robot-persona.md。入口用业务词语,真实执行后拟人表达;关窗计划仍运行。
- 账号/组织/对话/通知/本机及云资料:host-data-sdk.md;先声明能力,再调用真实SDK,分页读全,不编造接口。
- SDK/参数/鉴权/数据/MCP:host-sdk.md;复制同源SDK,读sdk/README.md;无顶层connect;宿主错误不伪装预览。
- 本地文件/导入/上传:local-files.md;原生桥选文件→会话;临时引用、持久素材、业务上传分开。
- 包/双语/交付:bridge-and-delivery.md;普通规范:plugin-development.md。
- package优先,复用宿主;本地结构:local-app-structure.md;Web/H5不限栈。
- 技能与页面联动:intelligent-interaction.md;写回/应用/追踪并验收。小任务ai.generate,复杂工作chat.send;见fast-ai.md。
- author真实;appCategory=office/game/robot/other;WEB/H5为接入方式。extensions保留未知字段、不放秘密。
- 数据遵从必选模式;契约与JSON。初始化版本化,不覆盖用户记录。
- 修改先读源码与 baseline.json;保留 key,提升 SemVer;资源完整收集,禁止省略未改文件。
- 最终磁盘运行 scripts/validate.cjs;失败修复重检。登记 .soduwork/delivery.json;走当前会话检查安装。静态检查不等于运行验收,不能虚构通过。
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.

