init: 毛孩子计划 小程序 + Go 后端 + 内嵌后台

- pets-fe: 微信原生小程序（首页/计划/记录/报告/社区/引导），
  服务端驱动、无假数据；弹层改用 scroll-view，打开时隐藏自定义 tabBar
- pets-be: Gin + GORM(MySQL, sundynix_ 前缀) + MinIO，统一响应/分页，
  微信 code2session 登录，provider-neutral AI(DeepSeek)，go:embed React 后台
- 修复：分段选择类型不匹配（字符串 vs 数字）导致选不中

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

pets-be/internal/ai/openai_compat.go

@@ -0,0 +1,122 @@
+package ai
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"strings"
+	"time"
+
+	"github.com/sundynix/pets-be/internal/config"
+)
+
+// openaiCompatProvider 走 OpenAI 兼容的 /chat/completions 协议。
+// DeepSeek、Qwen(dashscope 兼容模式)、Moonshot、Gemini(OpenAI 兼容端点) 等均可用。
+type openaiCompatProvider struct {
+	baseURL string
+	apiKey  string
+	model   string
+	client  *http.Client
+}
+
+func newOpenAICompatProvider(cfg config.AIConfig) *openaiCompatProvider {
+	timeout := time.Duration(cfg.TimeoutSec) * time.Second
+	if timeout <= 0 {
+		timeout = 30 * time.Second
+	}
+	return &openaiCompatProvider{
+		baseURL: strings.TrimRight(cfg.BaseURL, "/"),
+		apiKey:  cfg.APIKey,
+		model:   cfg.Model,
+		client:  &http.Client{Timeout: timeout},
+	}
+}
+
+func (p *openaiCompatProvider) Name() string { return "openai-compat:" + p.model }
+
+type chatReq struct {
+	Model          string      `json:"model"`
+	Messages       []chatMsg   `json:"messages"`
+	Temperature    float64     `json:"temperature,omitempty"`
+	MaxTokens      int         `json:"max_tokens,omitempty"`
+	ResponseFormat *respFormat `json:"response_format,omitempty"`
+}
+
+type chatMsg struct {
+	Role    string `json:"role"`
+	Content string `json:"content"`
+}
+
+type respFormat struct {
+	Type string `json:"type"`
+}
+
+type chatResp struct {
+	Choices []struct {
+		Message struct {
+			Content string `json:"content"`
+		} `json:"message"`
+	} `json:"choices"`
+	Error *struct {
+		Message string `json:"message"`
+	} `json:"error"`
+}
+
+func (p *openaiCompatProvider) Complete(system string, messages []Message, opts Options) (string, error) {
+	msgs := make([]chatMsg, 0, len(messages)+1)
+	if system != "" {
+		msgs = append(msgs, chatMsg{Role: "system", Content: system})
+	}
+	for _, m := range messages {
+		msgs = append(msgs, chatMsg{Role: m.Role, Content: m.Content})
+	}
+
+	body := chatReq{
+		Model:       p.model,
+		Messages:    msgs,
+		Temperature: opts.Temperature,
+		MaxTokens:   opts.MaxTokens,
+	}
+	if opts.JSON {
+		body.ResponseFormat = &respFormat{Type: "json_object"}
+	}
+	raw, _ := json.Marshal(body)
+
+	ctx, cancel := context.WithTimeout(context.Background(), p.client.Timeout)
+	defer cancel()
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, p.baseURL+"/chat/completions", bytes.NewReader(raw))
+	if err != nil {
+		return "", err
+	}
+	req.Header.Set("Content-Type", "application/json")
+	req.Header.Set("Authorization", "Bearer "+p.apiKey)
+
+	resp, err := p.client.Do(req)
+	if err != nil {
+		return "", err
+	}
+	defer resp.Body.Close()
+
+	data, _ := io.ReadAll(resp.Body)
+	var out chatResp
+	if err := json.Unmarshal(data, &out); err != nil {
+		return "", fmt.Errorf("ai 响应解析失败: %s", truncate(string(data), 200))
+	}
+	if out.Error != nil {
+		return "", fmt.Errorf("ai 供应商错误: %s", out.Error.Message)
+	}
+	if resp.StatusCode >= 400 || len(out.Choices) == 0 {
+		return "", fmt.Errorf("ai 调用失败(%d): %s", resp.StatusCode, truncate(string(data), 200))
+	}
+	return strings.TrimSpace(out.Choices[0].Message.Content), nil
+}
+
+func truncate(s string, n int) string {
+	if len(s) <= n {
+		return s
+	}
+	return s[:n]
+}