数据库迁移
- 作者仓库星标 11
- 作者更新于 实时读取
- 作者仓库 lovable-claude-code
- 领域
- 通用
- 兼容 Agent
-
- Claude Code
- Cursor
- Cline
- Codex
- Windsurf
- Gemini CLI
- +20
- 信任分
- 88 / 100 · 社区维护
- 作者 / 版本 / 许可
- @10K-Digital · 未声明 license
- Token 消耗评级
- 低消耗
- 接入复杂程度
- 即装即用
- 是否需要外部 API Key
- 不需要
- 兼容的系统
- 未声明(默认跨平台)
- 底层运行要求
- 无特殊要求
- 文件与系统权限
-
- 只读
- 允许写入 / 修改
- 读取环境变量
- 网络行为
- 仅限本地
- 安装命令数
- 26 条
档案由构建时根据 SKILL.md 与安装命令自动衍生,可能与作者实际意图存在差异。
需要注意: 未限定 allowed-tools,默认拥有全部工具权限。
---
name: lovable
description: | This skill enables Claude Code to work effectively with Lovable.dev projects while respecting…
category: 通用
runtime: 无特殊运行时
---
# lovable 输出预览
## PART A: 任务判断
- 适用问题:通用任务拆解、检查和交付。
- 输入要求:目标材料、限制条件、期望输出和验收方式。
- 证据边界:围绕“When to Use This Skill / Core Concept / What Syncs Automatically (GitHub → Lovable)”读取原文规则,不把推断写成作者承诺。
## PART B: 执行结果
- **01** 任务判断:确认你的需求是否属于通用任务拆解、检查和交付,并标出输入、限制和预期结果。
- **02** 执行计划:优先按“When to Use This Skill / Core Concept / What Syncs Automatically (GitHub → Lovable)”拆成步骤,说明每一步会读取什么、修改什么、产出什么。
- **03** 交付结果:给出可复制的命令、文件改动、检查清单或内容草稿,并说明如何继续迭代。
- **04** 风险边界:结合 读取文件、写入/修改文件、读取环境变量、主要在本地完成、通常不需要额外 API Key 给出执行前确认项。
## Running Rules
- 读取文件、写入/修改文件、读取环境变量;主要在本地完成;通常不需要额外 API Key。
- 先小样例验证,再放大到真实任务。
- 交付时同时给结果、检查口径和下一步迭代建议。 原文出现了 `/skills`、`/lovable` 这类斜杠命令;如果你的 Agent 支持命令触发,优先用命令开场,再补充目标和边界。
告诉 Agent 目标文件或材料、期望结果、不可改范围、是否允许联网或执行命令。本 Skill 的权限画像是:读取文件、写入/修改文件、读取环境变量。
先用一个小任务确认它会围绕“When to Use This Skill / Core Concept / What Syncs Automatically (GitHub → Lovable)”工作;涉及文件或命令时,先看 diff、日志、预览或测试结果。
检查最终产物是否包含明确结果、必要证据和下一步动作;如果输出泛泛而谈,就补充输入、边界和验收标准后重跑。
---
name: lovable
description: | This skill enables Claude Code to work effectively with Lovable.dev projects while respecting…
category: 通用
source: 10K-Digital/lovable-claude-code
---
# lovable
## 什么时候使用
- lovable 是一个通用扩展技能,按 SKILL 适合处理通用任务拆解、检查、交付和复盘,核心价值是把输入、判断、执行、验证和交付边界固定下来,避免 Agent 泛泛回答。 把任务拆成可执行、可检查、可继续迭代的步骤;通常不需要额外…
- 面向通用任务拆解、检查和交付,优先处理能明确输入、步骤和验收标准的工作。
## 需要提供什么
- 目标材料、目录范围、期望结果和不可改动内容。
- 是否允许联网、执行命令、读写文件或调用外部服务。
## 执行规则
- 围绕「When to Use This Skill / Core Concept / What Syncs Automatically (GitHub → Lovable)」组织步骤,不把推断写成作者事实。
- 读取文件、写入/修改文件、读取环境变量;主要在本地完成;通常不需要额外 API Key。
- 先跑小样例,确认结果可检查后再扩大任务范围。
## 输出要求
- 给出最终产物、关键证据、验证方式和下一步动作。
- 信息不足时标记 unknown,不编造命令、平台或依赖。 作者原文负责流程事实;仓库文件负责来源和命令;流狐只补充适用场景、限制和质量判断。
skill "lovable" {
输入层 -> 用户目标 + 目标文件 + 禁止范围 + 验收标准
上下文层 -> When to Use This Skill / Core Concept / What Syncs Automatically (GitHub → Lovable)
规则层 -> SKILL.md 触发条件 / 执行顺序 / 输出格式
运行层 -> 无特殊运行时 | 读取文件、写入/修改文件、读取环境变量 | 主要在本地完成
安全层 -> 通常不需要额外 API Key + 小任务验证 + diff / 日志复核
输出层 -> 可复制结果 + 检查清单 + 下一步迭代
} Lovable Integration Skill
This skill enables Claude Code to work effectively with Lovable.dev projects while respecting Lovable's deployment requirements.
When to Use This Skill
Activate when:
- User mentions "Lovable" or "lovable.dev"
- Project has
supabase/directory with Edge Functions - User asks to deploy edge functions
- User creates database migrations
- User asks about Lovable Cloud or backend deployment
- Project appears to be a Lovable project (React + Supabase structure)
Core Concept
Lovable uses two-way GitHub sync on the main branch only:
- Frontend code syncs automatically
- Backend operations (Edge Functions, migrations, RLS) require Lovable prompts
What Syncs Automatically (GitHub → Lovable)
✅ Edit freely and push to main:
src/- All React components, pages, hooks, utilspublic/- Static assets- Config files - vite.config.ts, tailwind.config.js, tsconfig.json
package.json- Dependenciessupabase/functions/*/index.ts- Edge Function code (not deployment)supabase/migrations/*.sql- Migration files (not application)
What Requires Lovable Deployment
⚠️ After editing, provide Lovable prompt:
| Change Type | Lovable Prompt |
|---|---|
| Edge Function code | "Deploy the [name] edge function" |
| All Edge Functions | "Deploy all edge functions" |
| New migration file | "Apply pending Supabase migrations" |
| New table needed | "Create a [name] table with columns: [list]" |
| RLS policy | "Enable RLS on [table] allowing [who] to [what]" |
| Storage bucket | "Create a [public/private] bucket called [name]" |
| Secret/env var | Manual: Cloud → Secrets → Add |
Response Format
When backend deployment is needed, always output:
📋 **LOVABLE PROMPT:**
> "[exact prompt to copy-paste]"
For destructive operations, add:
⚠️ **Warning**: [explanation of risk]
File Structure Reference
project/
├── src/ # ✅ Safe - auto-syncs
│ ├── components/
│ ├── pages/
│ ├── hooks/
│ ├── lib/
│ └── integrations/supabase/
│ ├── client.ts # ⚠️ Has Supabase URLs
│ └── types.ts
├── supabase/
│ ├── functions/ # ✅ Edit code, ⚠️ needs deploy
│ │ └── [function-name]/
│ │ └── index.ts
│ ├── migrations/ # ✅ Create files, ⚠️ needs apply
│ │ └── YYYYMMDDHHMMSS_*.sql
│ └── config.toml # ⚠️ Lovable Cloud manages
├── .env # Local only - Lovable ignores
└── CLAUDE.md # Project context
Backend Types
Lovable Cloud
- Backend managed entirely by Lovable
- No Supabase dashboard access
- All operations via Lovable prompts
- Secrets in Cloud → Secrets UI
Own Supabase
- Direct Supabase dashboard access
- Can use Supabase CLI:
supabase functions deploy - More flexibility but manual setup
Quick Prompts Reference
Edge Functions
"Deploy all edge functions"
"Deploy the send-email edge function"
"Create an edge function called [name] that [description]"
"Show logs for [name] edge function"
"The [name] edge function returns [error]. Fix it"
Database
"Create a [name] table with columns: id (uuid), name (text), created_at (timestamp)"
"Add a [column] column of type [type] to [table]"
"Add foreign key from [table1].[col] to [table2].id"
"Apply pending Supabase migrations"
RLS Policies
"Enable RLS on [table]"
"Add RLS policy on [table] allowing authenticated users to read all rows"
"Add RLS policy on [table] allowing users to only access their own rows"
Storage
"Create a public storage bucket called [name]"
"Create a private storage bucket called [name]"
"Allow authenticated users to upload to [bucket]"
Auth
"Enable Google authentication"
"Enable GitHub authentication"
"When user signs up, create row in profiles table"
Branch Rules
- Only
mainsyncs with Lovable - Feature branches don't deploy until merged
- Lovable syncs within 1-2 minutes of push
Yolo Mode - Automated Deployments (Beta)
When yolo_mode: on in CLAUDE.md, deployments are automated via browser automation:
How It Works
Instead of showing manual prompts, the yolo skill (/skills/yolo/SKILL.md) takes over:
- Automatically navigates to Lovable.dev
- Submits deployment prompts
- Monitors for success/failure
- Runs verification tests (if enabled)
- Reports deployment summary
When Yolo Mode Activates
- During
/lovable:deploy-edgecommand - During
/lovable:apply-migrationcommand - When
yolo_mode: onin CLAUDE.md
Configure Yolo Mode
/lovable:yolo on # Enable with testing
/lovable:yolo on --no-testing # Enable without testing
/lovable:yolo on --debug # Enable with verbose logs
/lovable:yolo off # Disable
Beta Status
⚠️ Yolo mode is in beta:
- Requires Claude in Chrome extension
- May have bugs or UI compatibility issues
- Always has manual fallback
- See
/skills/yolo/SKILL.mdfor details
Debugging Checklist
Frontend not updating?
- On
mainbranch? - Changes pushed?
- Wait 1-2 min
- On
Edge Function not working?
- Deployed via Lovable (or yolo mode)?
- Secrets set in Cloud UI?
- Check logs in Lovable
Database query failing?
- Migration applied (via Lovable or yolo mode)?
- RLS policies correct?
- Table exists?
Yolo mode not working?
- Check
yolo_mode: onin CLAUDE.md - Chrome extension installed?
- Logged into Lovable?
- See yolo skill for troubleshooting
- Check
先判断是否适合
作者设计意图
作者的方法与取舍
边界和复核