HappyCoding AI Coding

Facilitates full-cycle coding workflow in HappyCoding AI Coding: from requirement discussion, code edits, preview, to saving, publishing, and evidence docume...

Security Scan

The skill is an instruction-only workflow for a specific web platform and mostly coherent, but it implicitly relies on a logged-in account and references an API key endpoint without declaring required credentials — a mismatch that warrants caution.

Purpose & Capability

The name and description claim a full-cycle workflow for the '小畅AI Coding' platform and the SKILL.md only contains step-by-step instructions for using that platform (editing, previewing, saving, publishing). All required actions are consistent with that purpose.

Instruction Scope

Instructions include running JavaScript in the page (window.editorManager...), reading editor contents, saving/publishing projects, taking screenshots, and calling a multimodal image API. These are within the platform scope, but the skill expects to use the user's existing logged-in account and to '回传'publish links/screenshots — which are privacy-relevant behaviors the user should be aware of.

Install Mechanism

No install spec and no code files are present (instruction-only), so nothing is written to disk or installed. This is the lowest-risk install profile.

Credentials

The SKILL.md references an API endpoint that requires an x-api-key (placeholder {{your key}}) but the skill metadata declares no required environment variables or credentials. It also assumes a logged-in account '小畅openclaw' is available. The missing explicit declaration of credentials/session requirements is a proportionality mismatch and a transparency gap.

Persistence & Privilege

always is false and the skill does not request persistent system-level privileges or modify other skills. It instructs interacting with the web app and localStorage keys (localStorage.imageHistory), which is expected for a UI workflow; no unusual persistence or privilege escalation is requested.

What to consider before installing

This skill is an instruction-only guide for using a specific web coding platform and appears to do what it says, but note two things before using it: (1) it expects to use your existing logged-in account and to save/publish content — verify you are comfortable having the agent operate under that account and that published content won't leak secrets; (2) the SKILL.md references an image API requiring an x-api-key but the skill metadata does not declare any required credentials — confirm where that key comes from and only provide it if you trust the endpoint (https://ainet.aigcuniversity.com). If you plan to use this skill, ask the author to explicitly declare required credentials/session assumptions and to document where published links/screenshots will be sent or stored. If you are uncertain, test on a disposable account and avoid providing real API keys or private data.

xiaochang-ai-coding Skills（教学版）

xiaochang-ai-coding Skills（教学版）

用途：给"另外的AI"做教学时，复用这套在小畅AI Coding平台上的标准工作流。

0. Skill定位

• 名称：xiaochang-ai-coding
• 目标：在小畅AI Coding里完成从需求沟通、代码修改、预览验证、证据留档到发布的闭环。
• 适用场景：任何需要在"小畅AI Coding"中实现或迭代功能的任务。

1. 平台入口与会话

• 入口：https://aicoding.aigcuniversity.com/
• 账号：小畅openclaw（通常已登录，失效再处理）
• 建议先确认页面状态（项目列表/当前项目/编辑器是否可用）

2. 标准流程（最重要）

• 需求交流区先行：先在"需求交流区"描述需求，等平台生成修改建议。
• 自动建议优先：用"继续完善/添加更多功能/帮我美化"等按钮迭代。
• 必要时再手改代码：自动补丁无效或细节不到位时，进入代码区精修。
• 运行验证：运行后看预览区是否满足需求。
• 保存到项目：必须点左侧"我的项目"下方绿色保存（这是项目级提交）。
• 发布并回传链接：发布成功后记录分享链接与关键截图。

原则：优先保留"需求→修改→验证"的审计链，不要一上来就脱离平台流程硬改代码。

3. 代码区操作要点

3.1 读取当前代码

JavaScript
window.editorManager.editor.state.doc.toString()

3.2 批量覆盖代码（推荐）

• 先清空编辑器，再按3~4KB分块插入新代码。
• 局部replace可用，但整页覆盖更稳定。

3.3 运行与保存

• 运行按钮触发后，检查预览区。
• 最终一定要点项目级绿色"保存"，否则只在当前会话生效。

4. 发布规范

• 点击顶部发布
• 必须选择项目类型（常用"工具"）
• 确认名称后发布
• 拿到云端链接（如 https://t.istemedu.com/<slug>）并回传

5. 作画智能体（ID 10）API（项目常用）

• Endpoint：https://ainet.aigcuniversity.com/api/v1/tank/custom_gpt/10/execute
• Header：
• x-api-key: {{your key}}
• Content-Type: application/json
• 请求体核心：messages[] + stream:false
• 返回重点：result.images（Base64 PNG）
• 图生图：上传文件后转base64并写入init_image
• 错误处理：检查 response.ok、result.images 是否存在，异常时给UI状态提示

6. UI/交互约束（当前多模态助手项目）

• 视觉：柔和紫色风格
• 布局：左侧输入，右侧"生成记录区 + 示例图片区"分离
• 交互：
• Ctrl+Enter 快速生成
• 历史记录存 localStorage.imageHistory（最多10条）
• 空状态引导明确

7. 证据留档要求

• 保存页面快照/截图（结构与结果）
• 记录发布链接
• 记录关键改动点（例如：生成区与示例区分离）

8. 常见问题排查

• 编辑器无响应：刷新页面/重开代码面板/等待 editorManager 加载
• AI修改偏离：撤销后回到需求区重新约束
• 发布失败：检查项目类型是否已选、表单是否完整

9. 对外教学的一句话总结

先走"需求交流"，再做"必要代码修正"，最后完成"运行验证+项目级保存+发布回传"。
这套流程比单纯改代码更稳定，也更容易复盘教学过程。