feat(chat): support AGNES image/video models in AI chat - backend _generate_media (images/generations + async videos polling) and frontend MarkdownContent media rendering + model capability badges; nginx proxy_read_timeout raised to 600s for long media generation

backend/python/app/api/chat.py

@@ -12,6 +12,7 @@ import uuid
 from datetime import datetime, timezone, timedelta
 import re
 import json
+import asyncio
 import math
 import httpx
 import os
@@ -40,6 +41,108 @@ def cst_now():
     return datetime.now(CST_TZ)
 
 
+def _html_escape(s: str) -> str:
+    return (str(s).replace("&", "&amp;").replace("<", "&lt;").replace(">", "&gt;")
+            .replace('"', "&quot;").replace("'", "&#39;"))
+
+
+async def _generate_media(db, ai_model, prompt: str, user_id: str = None) -> str:
+    """图像/视频生成模型统一入口：按 capabilities 调用 AGNES 等媒体生成 API。
+    返回可直接在 AI 聊天中渲染的富文本(含 <img>/<video> 标签)。"""
+    provider = db.query(ModelProvider).filter(ModelProvider.id == ai_model.provider_id).first()
+    api_key = ai_model.api_key or (provider.api_key if provider else None)
+    if not api_key:
+        return "⚠️ 该媒体模型缺少 API Key，请在后台「大模型库」中配置。"
+
+    caps = [c.lower() for c in (ai_model.capabilities or [])]
+    headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
+    timeout = ai_model.timeout or 120
+
+    # ============ 图像生成 ============
+    if "image" in caps:
+        api_url = ai_model.endpoint
+        if not api_url:
+            api_url = f"{provider.api_base_url.rstrip('/')}/v1/images/generations" if provider else ""
+        if not api_url:
+            return "⚠️ 该图像模型未配置 Endpoint。"
+        payload = {
+            "model": ai_model.name,
+            "prompt": prompt,
+            "size": "1024x768",
+            "extra_body": {"response_format": "url"},
+        }
+        try:
+            async with httpx.AsyncClient(timeout=timeout) as client:
+                resp = await client.post(api_url, headers=headers, json=payload)
+            if resp.status_code != 200:
+                return f"⚠️ 图像生成失败 (HTTP {resp.status_code}): {resp.text[:200]}"
+            data = resp.json()
+            items = data.get("data") or []
+            if not items:
+                return f"⚠️ 图像生成失败：响应无图片数据 {str(data)[:200]}"
+            item = items[0]
+            url = item.get("url") or item.get("b64_json")
+            if not url:
+                return f"⚠️ 图像生成失败：未返回图片地址 {str(data)[:200]}"
+            if not url.startswith("http"):
+                url = f"data:image/png;base64,{url}"
+            return (f'<img class="md-img" src="{url}" alt="生成图片"/>\n\n'
+                    f'**提示词：** {_html_escape(prompt)}')
+        except Exception as e:
+            return f"⚠️ 图像生成异常：{str(e)[:200]}"
+
+    # ============ 视频生成 (异步任务 + 轮询) ============
+    if "video" in caps:
+        api_url = ai_model.endpoint
+        if not api_url:
+            api_url = f"{provider.api_base_url.rstrip('/')}/v1/videos" if provider else ""
+        if not api_url:
+            return "⚠️ 该视频模型未配置 Endpoint。"
+        base = api_url.split("/v1")[0] if "/v1" in api_url else api_url
+        payload = {
+            "model": ai_model.name,
+            "prompt": prompt,
+            "height": 768,
+            "width": 1152,
+            "num_frames": 121,
+            "frame_rate": 24,
+        }
+        try:
+            async with httpx.AsyncClient(timeout=timeout) as client:
+                resp = await client.post(api_url, headers=headers, json=payload)
+                if resp.status_code != 200:
+                    return f"⚠️ 视频生成失败 (HTTP {resp.status_code}): {resp.text[:200]}"
+                data = resp.json()
+                video_id = data.get("video_id") or data.get("task_id") or data.get("id")
+                if not video_id:
+                    return f"⚠️ 视频生成失败：未返回任务ID {str(data)[:200]}"
+                # 轮询结果(最多约 5 分钟)
+                result_url = None
+                err_msg = None
+                for _ in range(60):
+                    await asyncio.sleep(5)
+                    r2 = await client.get(f"{base}/agnesapi?video_id={video_id}", headers=headers)
+                    if r2.status_code != 200:
+                        continue
+                    rd = r2.json()
+                    st = rd.get("status")
+                    if st == "completed":
+                        result_url = rd.get("url")
+                        break
+                    if st == "failed":
+                        err_msg = rd.get("error")
+                        break
+                if result_url:
+                    return (f'<video class="md-video" controls src="{result_url}"></video>\n\n'
+                            f'**提示词：** {_html_escape(prompt)}')
+                return f"⚠️ 视频生成失败：{err_msg or '生成超时，请稍后在对话中重试'}"
+        except Exception as e:
+            return f"⚠️ 视频生成异常：{str(e)[:200]}"
+
+    return "⚠️ 未知的媒体模型类型。"
+
+
+
 # ============================================================
 # Request Models
 # ============================================================
@@ -881,7 +984,14 @@ async def send_message(
     if model_id:
         ai_model = db.query(AIModel).filter(AIModel.id == model_id).first()
 
-        if ai_model and ai_model.endpoint:
+        # ====== 图像/视频生成模型分支 ======
+        caps = [c.lower() for c in (ai_model.capabilities or [])] if ai_model else []
+        if ai_model and ("image" in caps or "video" in caps):
+            try:
+                ai_content = await _generate_media(db, ai_model, req.content, current_user.id)
+            except Exception as e:
+                ai_content = f"⚠️ 媒体生成失败：{str(e)[:200]}"
+        elif ai_model and ai_model.endpoint:
             provider = db.query(ModelProvider).filter(ModelProvider.id == ai_model.provider_id).first()
             api_key = ai_model.api_key or (provider.api_key if provider else None)
             api_url = ai_model.endpoint
@@ -1333,6 +1443,7 @@ def get_models(db: Session = Depends(get_db)):
             "display_name": getattr(m, "display_name", None) or m.name,
             "provider_id": m.provider_id,
             "provider_name": provider.name if provider else "",
+            "capabilities": m.capabilities or [],
             "status": m.status,
             "price_per_million_tokens": float(m.price_per_million_tokens or 0),
             "is_enabled": m.is_enabled,