接入成员画像 Dify 工作流并清理旧提取逻辑

- 新增 member_context 专用 DifyClient，统一兼容 completion 与 workflow 两种调用模式 - 将成员画像插件默认切换到 Dify workflow 模式，配置改用新的 workflow 应用与 workflows/run 接口 - 生成可直接导入 Dify 的成员画像工作流 DSL 文件，方便后台一键导入和发布 - 补充 Dify 工作流接入说明文档，明确输入字段、输出字段、发布步骤与插件消费方式 - 清理旧的单成员日摘要提取链路，日级画像统一收敛到群日批量提取路径，减少无效分支和历史残留 - 去除 member_context 内部多处旧 requests 直连调用，统一改为通过 DifyClient 调用 AI 服务 - 优化群日批量结果解析逻辑，只按 wxid 作为唯一主键识别成员，不再依赖昵称做唯一判断 - 新增按 wxid 的结果去重与完整度评分逻辑，遇到重复成员结果时优先保留字段更完整、置信度更高的一条 - 保留现有初始化、增量、周/月聚合与最终画像生成链路，同时剔除 workflow 接入后已无效或低价值的旧逻辑 - 为后续继续收紧 fallback 标记、增强后台质量诊断和优化工作流输出稳定性打下基础
2026-04-02 14:25:50 +08:00
parent 75dc3b289d
commit bfd0dbc15c
7 changed files with 673 additions and 185 deletions
--- a/plugins/member_context/dify_client.py
+++ b/plugins/member_context/dify_client.py
@@ -0,0 +1,116 @@
+# -*- coding: utf-8 -*-
+import json
+from typing import Dict, Optional
+
+import requests
+from loguru import logger
+
+
+class DifyClient:
+    """Dify completion/workflow 通用调用客户端"""
+
+    def __init__(self, api_config: Optional[Dict] = None):
+        api_config = api_config or {}
+        self.LOG = logger
+        self.enabled = bool(api_config.get("enable", api_config.get("enabled", False)))
+        self.base_url = (api_config.get("base_url") or "").rstrip("/")
+        self.api_key = api_config.get("api_key", "")
+        self.timeout = int(api_config.get("request_timeout", 60))
+        self.mode = str(api_config.get("mode", "completion")).strip().lower()
+        default_endpoint = "workflows/run" if self.mode == "workflow" else "completion-messages"
+        self.endpoint = str(api_config.get("endpoint", default_endpoint)).lstrip("/")
+        self.workflow_output_key = str(api_config.get("workflow_output_key", "text")).strip()
+
+    def is_available(self) -> bool:
+        return self.enabled and bool(self.base_url and self.api_key)
+
+    def run(self, prompt: str, user: str, inputs: Optional[Dict] = None,
+            tag: str = "") -> Optional[Dict]:
+        if not self.is_available():
+            return None
+
+        headers = {
+            "Authorization": f"Bearer {self.api_key}",
+            "Content-Type": "application/json",
+        }
+        payload_inputs = dict(inputs or {})
+        if self.mode == "completion":
+            payload_inputs.setdefault("query", prompt)
+        elif prompt and "query" not in payload_inputs:
+            payload_inputs["query"] = prompt
+
+        payload = {
+            "inputs": payload_inputs,
+            "response_mode": "blocking",
+            "user": user,
+        }
+        url = f"{self.base_url}/{self.endpoint}"
+        try:
+            self.LOG.info(f"[成员交互摘要][Dify] 发起请求: mode={self.mode}, endpoint={self.endpoint}, tag={tag}")
+            response = requests.post(url, headers=headers, json=payload, timeout=self.timeout)
+            response.raise_for_status()
+            data = response.json()
+            parsed = self._parse_response(data)
+            if parsed is not None:
+                return parsed
+            self.LOG.warning(
+                f"[成员交互摘要][Dify] 响应内容为空: mode={self.mode}, tag={tag}, "
+                f"response_preview={(response.text or '')[:300]}"
+            )
+            return None
+        except Exception as e:
+            self.LOG.warning(f"[成员交互摘要][Dify] 请求失败: mode={self.mode}, tag={tag}, error={e}")
+            return None
+
+    def _parse_response(self, data: Dict) -> Optional[Dict]:
+        if self.mode == "workflow":
+            return self._parse_workflow_response(data)
+        answer = data.get("answer", "")
+        usage = (data.get("metadata") or {}).get("usage", {}) or {}
+        return {
+            "text": str(answer or "").strip(),
+            "usage": usage,
+            "raw": data,
+        }
+
+    def _parse_workflow_response(self, data: Dict) -> Optional[Dict]:
+        payload = (data or {}).get("data", {}) or {}
+        outputs = payload.get("outputs", {}) or {}
+        text = ""
+
+        if self.workflow_output_key and outputs.get(self.workflow_output_key) is not None:
+            value = outputs.get(self.workflow_output_key)
+            text = self._stringify_output(value)
+        elif outputs.get("text") is not None:
+            text = self._stringify_output(outputs.get("text"))
+        elif outputs.get("answer") is not None:
+            text = self._stringify_output(outputs.get("answer"))
+        elif outputs.get("result_json") is not None:
+            text = self._stringify_output(outputs.get("result_json"))
+        elif outputs.get("result") is not None:
+            text = self._stringify_output(outputs.get("result"))
+        else:
+            for value in outputs.values():
+                text = self._stringify_output(value)
+                if text:
+                    break
+
+        usage = {
+            "total_tokens": payload.get("total_tokens"),
+            "latency": payload.get("elapsed_time"),
+        }
+        return {
+            "text": str(text or "").strip(),
+            "usage": usage,
+            "raw": data,
+        }
+
+    @staticmethod
+    def _stringify_output(value) -> str:
+        if value is None:
+            return ""
+        if isinstance(value, str):
+            return value.strip()
+        if isinstance(value, (dict, list)):
+            return json.dumps(value, ensure_ascii=False)
+        return str(value).strip()