skills/ → skills-dev(9), skills-req(10), skills-ops(4), skills-integration(8), skills-biz(4), skills-workflow(7) generate-marketplace.py 改为自动扫描所有 skills-* 目录。 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
65 lines
1.7 KiB
Markdown
65 lines
1.7 KiB
Markdown
# MCP 认证配置
|
||
|
||
## 环境选择原则
|
||
|
||
> 需求和任务的主数据源是**生产环境 (ai.pipexerp.com)**。
|
||
> 所有需求创建、任务创建、查询、更新操作必须在生产环境执行。
|
||
|
||
| 环境 | 用途 | API 地址 |
|
||
|------|------|----------|
|
||
| **生产环境** | 需求创建、任务创建、数据管理 | https://ai.pipexerp.com/api/v1 |
|
||
| 本地环境 | 代码开发调试、功能测试 | http://localhost:8080/api/v1 |
|
||
|
||
## MCP 配置
|
||
|
||
确保 `~/.claude/.mcp.json` 配置:
|
||
|
||
```json
|
||
{
|
||
"mcpServers": {
|
||
"ai-proj": {
|
||
"type": "stdio",
|
||
"command": "node",
|
||
"args": ["/path/to/mcp-task-bridge/dist/index.js"],
|
||
"env": {
|
||
"TASK_API_BASE": "https://ai.pipexerp.com/api/v1",
|
||
"TASK_API_TOKEN": "aiproj_pk_xxxxxxxx...",
|
||
"SYNC_REMOTE_API_BASE": "https://ai.pipexerp.com/api/v1",
|
||
"SYNC_REMOTE_API_KEY": "aiproj_pk_xxxxxxxx..."
|
||
}
|
||
}
|
||
}
|
||
}
|
||
```
|
||
|
||
## 认证方式
|
||
|
||
### PAT 秘钥(推荐)
|
||
|
||
- 格式: `aiproj_pk_` + 64 字符十六进制(共 74 字符)
|
||
- 优势: 长期有效,无需频繁更新
|
||
|
||
### dev-quick-login(备选)
|
||
|
||
仅在 PAT 秘钥不可用时使用:
|
||
|
||
```bash
|
||
curl -s -X POST "http://localhost:8080/api/v1/auth/dev-quick-login" \
|
||
-H "Content-Type: application/json" \
|
||
-d '{"username":"qiudl"}' | jq -r '.data.access_token'
|
||
```
|
||
|
||
> JWT token 有效期 24 小时,日常使用推荐 PAT 秘钥。
|
||
|
||
## 验证 PAT 秘钥
|
||
|
||
```bash
|
||
# 测试本地
|
||
curl -s -H "X-API-Key: aiproj_pk_xxx..." \
|
||
"http://localhost:8080/api/v1/tasks?limit=1" | jq '.success'
|
||
|
||
# 测试远程
|
||
curl -s -H "Authorization: Bearer aiproj_pk_xxx..." \
|
||
"https://ai.pipexerp.com/api/v1/tasks?limit=1" | jq '.success'
|
||
```
|