feat: 添加 ad_creator_agent 和 video_generator_agent
- ad_creator_agent: 多模态广告创意生成，支持 Gemini/GPT Image/DALL-E 图片生成和 Sora 视频生成 - video_generator_agent: 根据描述生成图片并拼接为视频 - 包含 Dockerfile、API 文档和测试 Made-with: Cursor
This commit is contained in:
zhanggangyong
@@ -0,0 +1,466 @@
|
||||
# Ad Creator Agent - API 文档
|
||||
|
||||
多模态广告创意生成 Agent，通过素材（文字描述/参考图片）生成广告图片或视频。
|
||||
|
||||
## 基本信息
|
||||
|
||||
| 项目 | 值 |
|
||||
|------|------|
|
||||
| 镜像 | `agnettaiji.azurecr.io/ai-agents/ad-creator-agent:latest` |
|
||||
| 端口 | `8000` |
|
||||
| 模板名 | `ad_creator_agent` |
|
||||
| 框架 | API (FastAPI) |
|
||||
|
||||
## 支持的模型
|
||||
|
||||
| 用途 | 模型 | 备注 |
|
||||
|------|------|------|
|
||||
| 图片生成（默认） | `taiji/gemini-3-pro-image-preview` | 支持参考图片输入 |
|
||||
| 图片生成 | `taiji/gpt-image-1` | OpenAI GPT Image |
|
||||
| 图片生成 | `taiji/gpt-image-1-mini` | 轻量版，速度更快 |
|
||||
| 图片生成 | `taiji/dall-e-3` | DALL-E 3 |
|
||||
| 文案生成 | `taiji/gpt-4o-mini` | 广告文案 + 图片 prompt |
|
||||
| 视频生成 | `taiji/sora-2` | Sora 视频生成 |
|
||||
|
||||
## 认证方式
|
||||
|
||||
所有写操作端点均需传入 API Key，支持以下两种方式：
|
||||
|
||||
```
|
||||
api-key: sk-xxx
|
||||
```
|
||||
|
||||
```
|
||||
Authorization: Bearer sk-xxx
|
||||
```
|
||||
|
||||
如果部署时配置了 `LLM_API_KEY` 环境变量，可省略请求头中的 Key。
|
||||
|
||||
## 环境变量
|
||||
|
||||
| 变量名 | 说明 | 默认值 |
|
||||
|--------|------|--------|
|
||||
| `LLM_API_KEY` | LiteLLM API Key | （必填或请求头传入） |
|
||||
| `LLM_BASE_URL` | LiteLLM Base URL | `https://litellm.graystone-fb459c5d.southeastasia.azurecontainerapps.io/v1` |
|
||||
| `DEFAULT_IMAGE_MODEL` | 默认图片模型 | `taiji/gemini-3-pro-image-preview` |
|
||||
| `DEFAULT_TEXT_MODEL` | 默认文案模型 | `taiji/gpt-4o-mini` |
|
||||
| `DEFAULT_VIDEO_MODEL` | 默认视频模型 | `taiji/sora-2` |
|
||||
| `SERVICE_PORT` | 服务端口 | `8000` |
|
||||
| `OUTPUT_DIR` | 文件输出目录 | `/app/outputs` |
|
||||
|
||||
---
|
||||
|
||||
## API 端点
|
||||
|
||||
### 1. 健康检查
|
||||
|
||||
**GET** `/health`
|
||||
|
||||
```bash
|
||||
curl http://<AGENT_URL>/health
|
||||
```
|
||||
|
||||
**响应示例:**
|
||||
|
||||
```json
|
||||
{
|
||||
"status": "healthy",
|
||||
"service": "Ad Creator Agent",
|
||||
"pod_name": "test-ad-creator",
|
||||
"models": {
|
||||
"image": "taiji/gemini-3-pro-image-preview",
|
||||
"text": "taiji/gpt-4o-mini",
|
||||
"video": "taiji/sora-2"
|
||||
},
|
||||
"callback_enabled": false,
|
||||
"timestamp": "2026-03-02T14:52:15.109589"
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 2. 生成广告图片
|
||||
|
||||
**POST** `/api/v1/generate-image`
|
||||
|
||||
通过文字描述生成广告图片，可指定模型、风格、宽高比等。
|
||||
|
||||
**请求体:**
|
||||
|
||||
| 字段 | 类型 | 必填 | 说明 |
|
||||
|------|------|------|------|
|
||||
| `prompt` | string | 是 | 广告图片描述/创意需求 |
|
||||
| `model` | string | 否 | 模型名称，默认 `taiji/gemini-3-pro-image-preview` |
|
||||
| `aspect_ratio` | string | 否 | 宽高比: `1:1`, `16:9`, `9:16`, `4:3`, `3:4`（Gemini） |
|
||||
| `size` | string | 否 | 图片尺寸（仅 GPT/DALL-E）: `1024x1024`, `1024x1792`, `1792x1024` |
|
||||
| `quality` | string | 否 | 质量: `low`, `medium`, `high`（默认 `high`） |
|
||||
| `style` | string | 否 | 广告风格: `modern`, `minimalist`, `luxury`, `playful`, `tech`, `vintage` |
|
||||
| `brand_name` | string | 否 | 品牌名称 |
|
||||
| `reference_image_b64` | string | 否 | 参考图片 base64（仅 Gemini 支持） |
|
||||
|
||||
**示例 - Gemini 生成:**
|
||||
|
||||
```bash
|
||||
curl -X POST http://<AGENT_URL>/api/v1/generate-image \
|
||||
-H "Content-Type: application/json" \
|
||||
-H "api-key: sk-xxx" \
|
||||
-d '{
|
||||
"prompt": "A premium headphone floating against dark gradient background with golden light accents",
|
||||
"aspect_ratio": "1:1",
|
||||
"quality": "high",
|
||||
"style": "luxury",
|
||||
"brand_name": "SoundElite"
|
||||
}'
|
||||
```
|
||||
|
||||
**示例 - GPT Image 生成:**
|
||||
|
||||
```bash
|
||||
curl -X POST http://<AGENT_URL>/api/v1/generate-image \
|
||||
-H "Content-Type: application/json" \
|
||||
-H "api-key: sk-xxx" \
|
||||
-d '{
|
||||
"prompt": "A vibrant Instagram ad for a coffee brand with warm morning light",
|
||||
"model": "taiji/gpt-image-1",
|
||||
"size": "1024x1024",
|
||||
"quality": "high"
|
||||
}'
|
||||
```
|
||||
|
||||
**响应示例:**
|
||||
|
||||
```json
|
||||
{
|
||||
"success": true,
|
||||
"file_path": "/app/outputs/images/ad_gemini_20260302_145310_209307.jpg",
|
||||
"filename": "ad_gemini_20260302_145310_209307.jpg",
|
||||
"url": "/api/v1/files/ad_gemini_20260302_145310_209307.jpg",
|
||||
"model": "taiji/gemini-3-pro-image-preview"
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 3. 上传参考图片并生成广告图
|
||||
|
||||
**POST** `/api/v1/generate-image-upload`
|
||||
|
||||
支持 `multipart/form-data` 上传参考图片，结合文字描述生成广告图。
|
||||
|
||||
**表单字段:**
|
||||
|
||||
| 字段 | 类型 | 必填 | 说明 |
|
||||
|------|------|------|------|
|
||||
| `prompt` | string | 是 | 广告图片描述 |
|
||||
| `reference_image` | file | 否 | 参考图片文件 |
|
||||
| `model` | string | 否 | 模型名称 |
|
||||
| `aspect_ratio` | string | 否 | 宽高比 |
|
||||
| `quality` | string | 否 | 质量 |
|
||||
| `style` | string | 否 | 广告风格 |
|
||||
| `brand_name` | string | 否 | 品牌名称 |
|
||||
|
||||
**示例:**
|
||||
|
||||
```bash
|
||||
curl -X POST http://<AGENT_URL>/api/v1/generate-image-upload \
|
||||
-H "api-key: sk-xxx" \
|
||||
-F "prompt=基于这张产品图，生成一张高端产品广告海报" \
|
||||
-F "reference_image=@product_photo.jpg" \
|
||||
-F "style=luxury" \
|
||||
-F "aspect_ratio=16:9"
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 4. 生成广告文案
|
||||
|
||||
**POST** `/api/v1/generate-copy`
|
||||
|
||||
根据产品信息，由 LLM 生成结构化广告文案（标题、正文、CTA、hashtags）以及用于图片生成的英文 prompt。
|
||||
|
||||
**请求体:**
|
||||
|
||||
| 字段 | 类型 | 必填 | 说明 |
|
||||
|------|------|------|------|
|
||||
| `product` | string | 是 | 产品/服务描述 |
|
||||
| `target_audience` | string | 否 | 目标受众 |
|
||||
| `tone` | string | 否 | 语气: `professional`, `casual`, `humorous`, `urgent`, `luxury` |
|
||||
| `platform` | string | 否 | 投放平台: `instagram`, `facebook`, `tiktok`, `billboard`, `general` |
|
||||
| `language` | string | 否 | 语言: `zh`, `en`, `ja`（默认 `zh`） |
|
||||
|
||||
**示例:**
|
||||
|
||||
```bash
|
||||
curl -X POST http://<AGENT_URL>/api/v1/generate-copy \
|
||||
-H "Content-Type: application/json" \
|
||||
-H "api-key: sk-xxx" \
|
||||
-d '{
|
||||
"product": "高端无线降噪耳机，主打沉浸式音乐体验",
|
||||
"target_audience": "音乐爱好者和商务人士",
|
||||
"tone": "luxury",
|
||||
"platform": "instagram",
|
||||
"language": "zh"
|
||||
}'
|
||||
```
|
||||
|
||||
**响应示例:**
|
||||
|
||||
```json
|
||||
{
|
||||
"success": true,
|
||||
"headline": "沉浸高端音质",
|
||||
"body_copy": "体验非凡音质，尽享音乐带来的宁静与专注...",
|
||||
"cta": "立即体验",
|
||||
"image_prompt": "A luxurious setting featuring a sleek wireless headphone...",
|
||||
"hashtags": ["#高端耳机", "#沉浸音乐", "#商务生活"]
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 5. 一键生成完整广告（文案 + 图片）
|
||||
|
||||
**POST** `/api/v1/generate-ad`
|
||||
|
||||
自动生成广告文案，并基于文案中的图片 prompt 自动生成配图。
|
||||
|
||||
**请求体:**
|
||||
|
||||
| 字段 | 类型 | 必填 | 说明 |
|
||||
|------|------|------|------|
|
||||
| `product` | string | 是 | 产品/服务描述 |
|
||||
| `image_model` | string | 否 | 图片生成模型 |
|
||||
| `aspect_ratio` | string | 否 | 宽高比 |
|
||||
| `style` | string | 否 | 广告风格 |
|
||||
| `brand_name` | string | 否 | 品牌名称 |
|
||||
| `target_audience` | string | 否 | 目标受众 |
|
||||
| `tone` | string | 否 | 语气 |
|
||||
| `platform` | string | 否 | 投放平台 |
|
||||
| `language` | string | 否 | 语言 |
|
||||
| `reference_image_b64` | string | 否 | 参考图片 base64 |
|
||||
|
||||
**示例:**
|
||||
|
||||
```bash
|
||||
curl -X POST http://<AGENT_URL>/api/v1/generate-ad \
|
||||
-H "Content-Type: application/json" \
|
||||
-H "api-key: sk-xxx" \
|
||||
-d '{
|
||||
"product": "新能源电动汽车，零排放、高续航、智能驾驶",
|
||||
"target_audience": "环保意识强的中产家庭",
|
||||
"tone": "professional",
|
||||
"platform": "facebook",
|
||||
"language": "zh",
|
||||
"style": "tech",
|
||||
"brand_name": "GreenDrive"
|
||||
}'
|
||||
```
|
||||
|
||||
**响应示例:**
|
||||
|
||||
```json
|
||||
{
|
||||
"success": true,
|
||||
"copy": {
|
||||
"success": true,
|
||||
"headline": "开启绿色出行新生活",
|
||||
"body_copy": "选择我们的新能源电动汽车...",
|
||||
"cta": "立即了解更多",
|
||||
"image_prompt": "A futuristic electric vehicle...",
|
||||
"hashtags": ["#新能源车", "#绿色出行", "#智能驾驶"]
|
||||
},
|
||||
"image": {
|
||||
"success": true,
|
||||
"filename": "ad_gemini_20260302_145504_262223.jpg",
|
||||
"url": "/api/v1/files/ad_gemini_20260302_145504_262223.jpg",
|
||||
"model": "taiji/gemini-3-pro-image-preview"
|
||||
},
|
||||
"timestamp": "2026-03-02T14:55:04.262223"
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 6. 生成广告视频
|
||||
|
||||
**POST** `/api/v1/generate-video`
|
||||
|
||||
使用 Sora 模型生成广告短视频。
|
||||
|
||||
**请求体:**
|
||||
|
||||
| 字段 | 类型 | 必填 | 说明 |
|
||||
|------|------|------|------|
|
||||
| `prompt` | string | 是 | 视频描述/创意需求 |
|
||||
| `model` | string | 否 | 视频模型（默认 `taiji/sora-2`） |
|
||||
| `aspect_ratio` | string | 否 | 宽高比: `16:9`, `9:16`, `1:1` |
|
||||
| `duration` | string | 否 | 视频时长秒数（默认 `5`） |
|
||||
|
||||
**示例:**
|
||||
|
||||
```bash
|
||||
curl -X POST http://<AGENT_URL>/api/v1/generate-video \
|
||||
-H "Content-Type: application/json" \
|
||||
-H "api-key: sk-xxx" \
|
||||
-d '{
|
||||
"prompt": "A sleek electric car driving through a futuristic city at sunset, cinematic style",
|
||||
"aspect_ratio": "16:9",
|
||||
"duration": "5"
|
||||
}'
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 7. 智能对话
|
||||
|
||||
**POST** `/chat`
|
||||
|
||||
与 AI 广告创意总监对话。系统会理解需求，自动决定是否生成图片。
|
||||
|
||||
**请求体:**
|
||||
|
||||
| 字段 | 类型 | 必填 | 说明 |
|
||||
|------|------|------|------|
|
||||
| `message` | string | 是 | 用户消息 |
|
||||
|
||||
**示例:**
|
||||
|
||||
```bash
|
||||
curl -X POST http://<AGENT_URL>/chat \
|
||||
-H "Content-Type: application/json" \
|
||||
-H "api-key: sk-xxx" \
|
||||
-d '{
|
||||
"message": "帮我为一款蓝牙音箱做一个抖音封面图，要有科技感"
|
||||
}'
|
||||
```
|
||||
|
||||
**响应示例:**
|
||||
|
||||
```json
|
||||
{
|
||||
"response": "为这款蓝牙音箱设计封面图的建议...",
|
||||
"image": {
|
||||
"success": true,
|
||||
"filename": "ad_gemini_20260302_145539_866923.jpg",
|
||||
"url": "/api/v1/files/ad_gemini_20260302_145539_866923.jpg",
|
||||
"model": "taiji/gemini-3-pro-image-preview"
|
||||
},
|
||||
"timestamp": "2026-03-02T14:55:39.866923"
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 8. 下载生成的文件
|
||||
|
||||
**GET** `/api/v1/files/{filename}`
|
||||
|
||||
```bash
|
||||
curl -O http://<AGENT_URL>/api/v1/files/ad_gemini_20260302_145310_209307.jpg
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 9. 列出已生成的文件
|
||||
|
||||
**GET** `/api/v1/list-files?file_type=all`
|
||||
|
||||
参数 `file_type` 可选值: `all`, `image`, `video`
|
||||
|
||||
```bash
|
||||
curl http://<AGENT_URL>/api/v1/list-files
|
||||
```
|
||||
|
||||
**响应示例:**
|
||||
|
||||
```json
|
||||
{
|
||||
"images": [
|
||||
{
|
||||
"filename": "ad_gemini_20260302_145539_866923.jpg",
|
||||
"url": "/api/v1/files/ad_gemini_20260302_145539_866923.jpg",
|
||||
"size_bytes": 589722,
|
||||
"created_at": "2026-03-02T14:55:39.865520"
|
||||
}
|
||||
],
|
||||
"videos": []
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 10. 清理旧文件
|
||||
|
||||
**POST** `/api/v1/cleanup?max_age_hours=24`
|
||||
|
||||
删除超过指定时间的旧文件。
|
||||
|
||||
```bash
|
||||
curl -X POST "http://<AGENT_URL>/api/v1/cleanup?max_age_hours=24"
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
### 11. 状态查看
|
||||
|
||||
**GET** `/status`
|
||||
|
||||
```bash
|
||||
curl http://<AGENT_URL>/status
|
||||
```
|
||||
|
||||
**响应示例:**
|
||||
|
||||
```json
|
||||
{
|
||||
"status": "running",
|
||||
"pod_name": "test-ad-creator",
|
||||
"generated_images": 4,
|
||||
"generated_videos": 0,
|
||||
"timestamp": "2026-03-02T15:01:43.636444"
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 通过 Agent Manager 部署
|
||||
|
||||
### 1. 注册模板
|
||||
|
||||
```bash
|
||||
curl -X POST http://20.212.121.126/templates/create \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"name": "ad_creator_agent",
|
||||
"display_name": "Ad Creator Agent",
|
||||
"description": "多模态广告创意生成 Agent",
|
||||
"image": "agnettaiji.azurecr.io/ai-agents/ad-creator-agent:latest",
|
||||
"port": 8000,
|
||||
"agent_type": "platform",
|
||||
"agent_framework": "api",
|
||||
"env_requirements": {
|
||||
"LLM_API_KEY": "LiteLLM API Key"
|
||||
}
|
||||
}'
|
||||
```
|
||||
|
||||
### 2. 创建实例
|
||||
|
||||
```bash
|
||||
curl -X POST http://20.212.121.126/agents \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"name": "my-ad-creator",
|
||||
"template": "ad_creator_agent",
|
||||
"config": { "user_id": "your-user-id" },
|
||||
"env": {
|
||||
"LLM_API_KEY": "sk-your-api-key"
|
||||
}
|
||||
}'
|
||||
```
|
||||
|
||||
### 3. 删除实例
|
||||
|
||||
```bash
|
||||
curl -X DELETE http://20.212.121.126/agents/my-ad-creator
|
||||
```
|
||||
@@ -0,0 +1,35 @@
|
||||
FROM python:3.11-slim
|
||||
|
||||
WORKDIR /app
|
||||
|
||||
RUN apt-get update && apt-get install -y \
|
||||
curl \
|
||||
&& rm -rf /var/lib/apt/lists/*
|
||||
|
||||
RUN pip install --no-cache-dir \
|
||||
fastapi==0.109.0 \
|
||||
uvicorn[standard]==0.27.0 \
|
||||
pydantic==2.5.3 \
|
||||
aiohttp>=3.9.0 \
|
||||
python-multipart>=0.0.6
|
||||
|
||||
COPY common/agent_callback_utils.py /app/common/
|
||||
RUN touch /app/common/__init__.py
|
||||
|
||||
COPY agents/ad_creator_agent/ad_creator_agent.py /app/
|
||||
|
||||
ENV PYTHONUNBUFFERED=1
|
||||
ENV SERVICE_HOST=0.0.0.0
|
||||
ENV SERVICE_PORT=8000
|
||||
ENV OUTPUT_DIR=/app/outputs
|
||||
|
||||
ENV AGENT_CALLBACK_URL=http://mcp-server:8002/api/v1/billing/agent-callback
|
||||
|
||||
RUN mkdir -p /app/outputs/images /app/outputs/videos
|
||||
|
||||
HEALTHCHECK --interval=30s --timeout=10s --start-period=20s --retries=3 \
|
||||
CMD python3 -c "import urllib.request; urllib.request.urlopen('http://localhost:8000/health').read()" || exit 1
|
||||
|
||||
EXPOSE 8000
|
||||
|
||||
CMD ["python3", "-u", "ad_creator_agent.py"]
|
||||
@@ -0,0 +1,793 @@
|
||||
"""
|
||||
Ad Creator Agent - 多模态广告创意生成 Agent
|
||||
通过素材（文字描述/参考图片）生成广告图片或视频
|
||||
支持模型：Gemini 3 Pro Image / GPT Image 1 / DALL-E 3 / Sora 2
|
||||
"""
|
||||
import os
|
||||
import sys
|
||||
import re
|
||||
import uuid
|
||||
import json
|
||||
import base64
|
||||
import logging
|
||||
import aiohttp
|
||||
from typing import Optional, List, Dict, Any
|
||||
from datetime import datetime
|
||||
from pathlib import Path
|
||||
from enum import Enum
|
||||
|
||||
from fastapi import FastAPI, HTTPException, Header, Depends, UploadFile, File, Form, Request
|
||||
from fastapi.middleware.cors import CORSMiddleware
|
||||
from fastapi.responses import FileResponse, JSONResponse
|
||||
from pydantic import BaseModel, Field
|
||||
import uvicorn
|
||||
|
||||
sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
||||
|
||||
try:
|
||||
from common.agent_callback_utils import AgentCallbackHandler, CallbackContextManager
|
||||
CALLBACK_ENABLED = True
|
||||
except ImportError:
|
||||
CALLBACK_ENABLED = False
|
||||
AgentCallbackHandler = None
|
||||
CallbackContextManager = None
|
||||
|
||||
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
# ==================== 环境变量 ====================
|
||||
|
||||
SERVICE_HOST = os.getenv("SERVICE_HOST", "0.0.0.0")
|
||||
SERVICE_PORT = int(os.getenv("SERVICE_PORT", "8000"))
|
||||
POD_NAME = os.getenv("POD_NAME", "ad-creator-agent")
|
||||
USER_ID = os.getenv("USER_ID", "")
|
||||
OUTPUT_DIR = os.getenv("OUTPUT_DIR", "/app/outputs")
|
||||
|
||||
LLM_BASE_URL = os.getenv(
|
||||
"LLM_BASE_URL",
|
||||
"https://litellm.graystone-fb459c5d.southeastasia.azurecontainerapps.io/v1"
|
||||
)
|
||||
LLM_API_KEY = os.getenv("LLM_API_KEY", "")
|
||||
|
||||
DEFAULT_IMAGE_MODEL = os.getenv("DEFAULT_IMAGE_MODEL", "taiji/gemini-3-pro-image-preview")
|
||||
DEFAULT_TEXT_MODEL = os.getenv("DEFAULT_TEXT_MODEL", "taiji/gpt-4o-mini")
|
||||
DEFAULT_VIDEO_MODEL = os.getenv("DEFAULT_VIDEO_MODEL", "taiji/sora-2")
|
||||
|
||||
Path(OUTPUT_DIR).mkdir(parents=True, exist_ok=True)
|
||||
Path(f"{OUTPUT_DIR}/images").mkdir(parents=True, exist_ok=True)
|
||||
Path(f"{OUTPUT_DIR}/videos").mkdir(parents=True, exist_ok=True)
|
||||
|
||||
|
||||
# ==================== 模型枚举 ====================
|
||||
|
||||
class ImageModel(str, Enum):
|
||||
GEMINI = "taiji/gemini-3-pro-image-preview"
|
||||
GPT_IMAGE = "taiji/gpt-image-1"
|
||||
GPT_IMAGE_MINI = "taiji/gpt-image-1-mini"
|
||||
DALLE3 = "taiji/dall-e-3"
|
||||
|
||||
|
||||
class AspectRatio(str, Enum):
|
||||
SQUARE = "1:1"
|
||||
LANDSCAPE = "16:9"
|
||||
PORTRAIT = "9:16"
|
||||
WIDE = "4:3"
|
||||
TALL = "3:4"
|
||||
|
||||
|
||||
# ==================== 请求/响应模型 ====================
|
||||
|
||||
class GenerateImageRequest(BaseModel):
|
||||
prompt: str = Field(..., description="广告图片描述/创意需求")
|
||||
model: Optional[str] = Field(None, description="图片生成模型，默认 gemini-3-pro-image-preview")
|
||||
aspect_ratio: Optional[str] = Field("1:1", description="宽高比: 1:1, 16:9, 9:16, 4:3, 3:4")
|
||||
size: Optional[str] = Field("1024x1024", description="图片尺寸（仅 GPT/DALL-E）")
|
||||
quality: Optional[str] = Field("high", description="质量: low, medium, high")
|
||||
style: Optional[str] = Field(None, description="广告风格: modern, minimalist, luxury, playful, tech, vintage")
|
||||
brand_name: Optional[str] = Field(None, description="品牌名称")
|
||||
reference_image_b64: Optional[str] = Field(None, description="参考图片 base64 (用于风格参考或产品素材)")
|
||||
user_id: Optional[str] = Field(None, description="用户ID")
|
||||
|
||||
|
||||
class GenerateAdCopyRequest(BaseModel):
|
||||
product: str = Field(..., description="产品/服务描述")
|
||||
target_audience: Optional[str] = Field(None, description="目标受众")
|
||||
tone: Optional[str] = Field("professional", description="语气: professional, casual, humorous, urgent, luxury")
|
||||
platform: Optional[str] = Field("general", description="投放平台: instagram, facebook, tiktok, billboard, general")
|
||||
language: Optional[str] = Field("zh", description="语言: zh, en, ja")
|
||||
user_id: Optional[str] = Field(None, description="用户ID")
|
||||
|
||||
|
||||
class GenerateAdRequest(BaseModel):
|
||||
"""完整广告生成（文案 + 图片）"""
|
||||
product: str = Field(..., description="产品/服务描述")
|
||||
image_model: Optional[str] = Field(None, description="图片生成模型")
|
||||
aspect_ratio: Optional[str] = Field("1:1", description="宽高比")
|
||||
style: Optional[str] = Field(None, description="广告风格")
|
||||
brand_name: Optional[str] = Field(None, description="品牌名称")
|
||||
target_audience: Optional[str] = Field(None, description="目标受众")
|
||||
tone: Optional[str] = Field("professional", description="语气")
|
||||
platform: Optional[str] = Field("general", description="投放平台")
|
||||
language: Optional[str] = Field("zh", description="语言")
|
||||
reference_image_b64: Optional[str] = Field(None, description="参考图片 base64")
|
||||
user_id: Optional[str] = Field(None, description="用户ID")
|
||||
|
||||
|
||||
class GenerateVideoRequest(BaseModel):
|
||||
prompt: str = Field(..., description="视频描述/创意需求")
|
||||
model: Optional[str] = Field(None, description="视频模型, 默认 sora-2")
|
||||
aspect_ratio: Optional[str] = Field("16:9", description="宽高比")
|
||||
duration: Optional[str] = Field("5", description="视频时长秒数")
|
||||
user_id: Optional[str] = Field(None, description="用户ID")
|
||||
|
||||
|
||||
class ChatRequest(BaseModel):
|
||||
message: str = Field(..., description="用户消息")
|
||||
user_id: Optional[str] = Field(None, description="用户ID")
|
||||
|
||||
|
||||
# ==================== FastAPI ====================
|
||||
|
||||
app = FastAPI(
|
||||
title="Ad Creator Agent",
|
||||
description="多模态广告创意生成 - 通过素材生成广告图片或视频（支持 Gemini / GPT Image / DALL-E / Sora）",
|
||||
version="1.0.0"
|
||||
)
|
||||
|
||||
app.add_middleware(
|
||||
CORSMiddleware,
|
||||
allow_origins=["*"],
|
||||
allow_credentials=True,
|
||||
allow_methods=["*"],
|
||||
allow_headers=["*"],
|
||||
)
|
||||
|
||||
callback_handler: Optional[Any] = None
|
||||
|
||||
|
||||
# ==================== 依赖 ====================
|
||||
|
||||
async def get_api_key(
|
||||
api_key: Optional[str] = Header(None, alias="api-key"),
|
||||
authorization: Optional[str] = Header(None),
|
||||
) -> str:
|
||||
if api_key and api_key.strip():
|
||||
return api_key.strip()
|
||||
if authorization:
|
||||
key = authorization[7:].strip() if authorization.startswith("Bearer ") else authorization.strip()
|
||||
if key:
|
||||
return key
|
||||
if LLM_API_KEY:
|
||||
return LLM_API_KEY
|
||||
raise HTTPException(status_code=401, detail="请在请求头提供 api-key 或 Authorization Bearer token")
|
||||
|
||||
|
||||
# ==================== 生命周期 ====================
|
||||
|
||||
@app.on_event("startup")
|
||||
async def startup_event():
|
||||
global callback_handler
|
||||
if CALLBACK_ENABLED:
|
||||
callback_handler = AgentCallbackHandler(agent_name=POD_NAME, user_id=USER_ID)
|
||||
logger.info(f"回调处理器已初始化: {callback_handler.callback_url}")
|
||||
else:
|
||||
logger.warning("回调模块未加载")
|
||||
logger.info(f"Ad Creator Agent 启动: port={SERVICE_PORT}, output={OUTPUT_DIR}")
|
||||
logger.info(f"默认模型: image={DEFAULT_IMAGE_MODEL}, text={DEFAULT_TEXT_MODEL}, video={DEFAULT_VIDEO_MODEL}")
|
||||
|
||||
|
||||
# ==================== 核心：图片生成 ====================
|
||||
|
||||
async def generate_image_gemini(
|
||||
prompt: str,
|
||||
api_key: str,
|
||||
aspect_ratio: str = "1:1",
|
||||
quality: str = "high",
|
||||
reference_image_b64: str = None,
|
||||
) -> dict:
|
||||
"""通过 Gemini chat/completions 生成图片，返回 {success, file_path, filename, url}"""
|
||||
|
||||
quality_map = {"low": "1K", "medium": "1K", "high": "2K"}
|
||||
image_size = quality_map.get(quality, "2K")
|
||||
|
||||
messages_content: Any
|
||||
if reference_image_b64:
|
||||
messages_content = [
|
||||
{
|
||||
"type": "image_url",
|
||||
"image_url": {"url": f"data:image/png;base64,{reference_image_b64}"},
|
||||
},
|
||||
{"type": "text", "text": prompt},
|
||||
]
|
||||
else:
|
||||
messages_content = prompt
|
||||
|
||||
payload = {
|
||||
"model": DEFAULT_IMAGE_MODEL,
|
||||
"stream": False,
|
||||
"messages": [{"role": "user", "content": messages_content}],
|
||||
"extra_body": {
|
||||
"google": {
|
||||
"image_config": {
|
||||
"aspect_ratio": aspect_ratio,
|
||||
"image_size": image_size,
|
||||
}
|
||||
}
|
||||
},
|
||||
}
|
||||
|
||||
headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
|
||||
base = LLM_BASE_URL.rstrip("/")
|
||||
url = f"{base}/chat/completions"
|
||||
|
||||
async with aiohttp.ClientSession() as session:
|
||||
async with session.post(url, json=payload, headers=headers, timeout=aiohttp.ClientTimeout(total=120)) as resp:
|
||||
if resp.status != 200:
|
||||
text = await resp.text()
|
||||
return {"success": False, "error": f"Gemini API error {resp.status}: {text[:500]}"}
|
||||
|
||||
data = await resp.json()
|
||||
content = data.get("choices", [{}])[0].get("message", {}).get("content", "")
|
||||
|
||||
match = re.search(r"data:image/([\w+]+);base64,([A-Za-z0-9+/=\s]+)", content)
|
||||
if not match:
|
||||
return {"success": False, "error": "Gemini 未返回图片数据", "text_response": content[:500]}
|
||||
|
||||
img_format = match.group(1).replace("+xml", "")
|
||||
ext = "jpg" if img_format == "jpeg" else img_format
|
||||
b64_data = match.group(2).replace("\n", "").replace(" ", "")
|
||||
|
||||
ts = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
|
||||
filename = f"ad_gemini_{ts}.{ext}"
|
||||
file_path = os.path.join(OUTPUT_DIR, "images", filename)
|
||||
|
||||
with open(file_path, "wb") as f:
|
||||
f.write(base64.b64decode(b64_data))
|
||||
|
||||
logger.info(f"Gemini 图片已生成: {file_path} ({os.path.getsize(file_path)} bytes)")
|
||||
return {
|
||||
"success": True,
|
||||
"file_path": file_path,
|
||||
"filename": filename,
|
||||
"url": f"/api/v1/files/{filename}",
|
||||
"model": DEFAULT_IMAGE_MODEL,
|
||||
}
|
||||
|
||||
|
||||
async def generate_image_openai(
|
||||
prompt: str,
|
||||
api_key: str,
|
||||
model: str = "taiji/gpt-image-1",
|
||||
size: str = "1024x1024",
|
||||
quality: str = "high",
|
||||
) -> dict:
|
||||
"""通过 OpenAI images/generations 接口生成图片（GPT Image / DALL-E）"""
|
||||
|
||||
quality_map = {"low": "low", "medium": "medium", "high": "high"}
|
||||
payload = {
|
||||
"model": model,
|
||||
"prompt": prompt,
|
||||
"n": 1,
|
||||
"size": size,
|
||||
"quality": quality_map.get(quality, "high"),
|
||||
}
|
||||
|
||||
headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
|
||||
base = LLM_BASE_URL.rstrip("/")
|
||||
url = f"{base}/images/generations"
|
||||
|
||||
async with aiohttp.ClientSession() as session:
|
||||
async with session.post(url, json=payload, headers=headers, timeout=aiohttp.ClientTimeout(total=120)) as resp:
|
||||
if resp.status != 200:
|
||||
text = await resp.text()
|
||||
return {"success": False, "error": f"OpenAI API error {resp.status}: {text[:500]}"}
|
||||
|
||||
data = await resp.json()
|
||||
items = data.get("data", [])
|
||||
if not items:
|
||||
return {"success": False, "error": "API 未返回图片数据"}
|
||||
|
||||
item = items[0]
|
||||
b64_data = item.get("b64_json")
|
||||
image_url = item.get("url")
|
||||
|
||||
ts = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
|
||||
model_tag = model.split("/")[-1].replace("-", "")
|
||||
filename = f"ad_{model_tag}_{ts}.png"
|
||||
file_path = os.path.join(OUTPUT_DIR, "images", filename)
|
||||
|
||||
if b64_data:
|
||||
with open(file_path, "wb") as f:
|
||||
f.write(base64.b64decode(b64_data))
|
||||
elif image_url:
|
||||
async with session.get(image_url, timeout=aiohttp.ClientTimeout(total=30)) as dl_resp:
|
||||
if dl_resp.status != 200:
|
||||
return {"success": False, "error": f"下载图片失败: HTTP {dl_resp.status}"}
|
||||
with open(file_path, "wb") as f:
|
||||
f.write(await dl_resp.read())
|
||||
else:
|
||||
return {"success": False, "error": "API 响应中无图片数据"}
|
||||
|
||||
logger.info(f"OpenAI 图片已生成: {file_path} ({os.path.getsize(file_path)} bytes)")
|
||||
return {
|
||||
"success": True,
|
||||
"file_path": file_path,
|
||||
"filename": filename,
|
||||
"url": f"/api/v1/files/{filename}",
|
||||
"model": model,
|
||||
}
|
||||
|
||||
|
||||
async def generate_image_dispatch(
|
||||
prompt: str,
|
||||
api_key: str,
|
||||
model: str = None,
|
||||
aspect_ratio: str = "1:1",
|
||||
size: str = "1024x1024",
|
||||
quality: str = "high",
|
||||
reference_image_b64: str = None,
|
||||
) -> dict:
|
||||
"""根据模型分发到对应生成函数"""
|
||||
model = model or DEFAULT_IMAGE_MODEL
|
||||
|
||||
if "gemini" in model.lower():
|
||||
return await generate_image_gemini(
|
||||
prompt=prompt,
|
||||
api_key=api_key,
|
||||
aspect_ratio=aspect_ratio,
|
||||
quality=quality,
|
||||
reference_image_b64=reference_image_b64,
|
||||
)
|
||||
else:
|
||||
return await generate_image_openai(
|
||||
prompt=prompt,
|
||||
api_key=api_key,
|
||||
model=model,
|
||||
size=size,
|
||||
quality=quality,
|
||||
)
|
||||
|
||||
|
||||
# ==================== 核心：文案生成 ====================
|
||||
|
||||
async def generate_ad_copy(
|
||||
product: str,
|
||||
api_key: str,
|
||||
target_audience: str = None,
|
||||
tone: str = "professional",
|
||||
platform: str = "general",
|
||||
language: str = "zh",
|
||||
) -> dict:
|
||||
"""用 LLM 生成广告文案和图片 prompt"""
|
||||
|
||||
lang_map = {"zh": "中文", "en": "English", "ja": "日本語"}
|
||||
lang_label = lang_map.get(language, language)
|
||||
|
||||
system_prompt = f"""你是一位顶尖的广告创意总监。请根据产品信息生成广告创意方案。
|
||||
回复必须使用 {lang_label}。
|
||||
回复格式必须严格为以下 JSON（不包含 markdown 代码块标记）:
|
||||
{{
|
||||
"headline": "广告标题（10字以内）",
|
||||
"body_copy": "广告正文（30-60字）",
|
||||
"cta": "行动号召按钮文字",
|
||||
"image_prompt": "用于 AI 生成广告配图的英文详细描述（100-200 words, 描述画面构图、色调、元素）",
|
||||
"hashtags": ["标签1", "标签2", "标签3"]
|
||||
}}"""
|
||||
|
||||
user_msg = f"产品/服务: {product}"
|
||||
if target_audience:
|
||||
user_msg += f"\n目标受众: {target_audience}"
|
||||
user_msg += f"\n语气风格: {tone}\n投放平台: {platform}"
|
||||
|
||||
payload = {
|
||||
"model": DEFAULT_TEXT_MODEL,
|
||||
"messages": [
|
||||
{"role": "system", "content": system_prompt},
|
||||
{"role": "user", "content": user_msg},
|
||||
],
|
||||
"max_tokens": 800,
|
||||
"temperature": 0.8,
|
||||
}
|
||||
|
||||
headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
|
||||
base = LLM_BASE_URL.rstrip("/")
|
||||
url = f"{base}/chat/completions"
|
||||
|
||||
async with aiohttp.ClientSession() as session:
|
||||
async with session.post(url, json=payload, headers=headers, timeout=aiohttp.ClientTimeout(total=30)) as resp:
|
||||
if resp.status != 200:
|
||||
text = await resp.text()
|
||||
return {"success": False, "error": f"LLM API error {resp.status}: {text[:300]}"}
|
||||
data = await resp.json()
|
||||
content = data.get("choices", [{}])[0].get("message", {}).get("content", "")
|
||||
|
||||
try:
|
||||
cleaned = content.strip()
|
||||
if cleaned.startswith("```"):
|
||||
cleaned = re.sub(r"^```\w*\n?", "", cleaned)
|
||||
cleaned = re.sub(r"\n?```$", "", cleaned)
|
||||
copy_data = json.loads(cleaned)
|
||||
except json.JSONDecodeError:
|
||||
return {
|
||||
"success": True,
|
||||
"raw_text": content,
|
||||
"headline": "",
|
||||
"body_copy": content[:200],
|
||||
"cta": "",
|
||||
"image_prompt": "",
|
||||
"hashtags": [],
|
||||
}
|
||||
|
||||
return {
|
||||
"success": True,
|
||||
"headline": copy_data.get("headline", ""),
|
||||
"body_copy": copy_data.get("body_copy", ""),
|
||||
"cta": copy_data.get("cta", ""),
|
||||
"image_prompt": copy_data.get("image_prompt", ""),
|
||||
"hashtags": copy_data.get("hashtags", []),
|
||||
}
|
||||
|
||||
|
||||
# ==================== 核心：视频生成 ====================
|
||||
|
||||
async def generate_video_sora(
|
||||
prompt: str,
|
||||
api_key: str,
|
||||
model: str = None,
|
||||
aspect_ratio: str = "16:9",
|
||||
duration: str = "5",
|
||||
) -> dict:
|
||||
"""通过 Sora 生成视频（OpenAI 兼容接口）"""
|
||||
model = model or DEFAULT_VIDEO_MODEL
|
||||
|
||||
size_map = {
|
||||
"16:9": "1920x1080",
|
||||
"9:16": "1080x1920",
|
||||
"1:1": "1080x1080",
|
||||
}
|
||||
size = size_map.get(aspect_ratio, "1920x1080")
|
||||
|
||||
payload = {
|
||||
"model": model,
|
||||
"input": prompt,
|
||||
"size": size,
|
||||
"duration": int(duration),
|
||||
"n": 1,
|
||||
}
|
||||
|
||||
headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
|
||||
base = LLM_BASE_URL.rstrip("/")
|
||||
url = f"{base}/videos/generations"
|
||||
|
||||
async with aiohttp.ClientSession() as session:
|
||||
async with session.post(url, json=payload, headers=headers, timeout=aiohttp.ClientTimeout(total=300)) as resp:
|
||||
if resp.status != 200:
|
||||
text = await resp.text()
|
||||
return {"success": False, "error": f"Sora API error {resp.status}: {text[:500]}"}
|
||||
|
||||
data = await resp.json()
|
||||
items = data.get("data", [])
|
||||
if not items:
|
||||
return {"success": False, "error": "Sora 未返回视频数据"}
|
||||
|
||||
video_url = items[0].get("url")
|
||||
b64_data = items[0].get("b64_json")
|
||||
|
||||
ts = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
|
||||
filename = f"ad_video_{ts}.mp4"
|
||||
file_path = os.path.join(OUTPUT_DIR, "videos", filename)
|
||||
|
||||
if b64_data:
|
||||
with open(file_path, "wb") as f:
|
||||
f.write(base64.b64decode(b64_data))
|
||||
elif video_url:
|
||||
async with session.get(video_url, timeout=aiohttp.ClientTimeout(total=120)) as dl_resp:
|
||||
if dl_resp.status != 200:
|
||||
return {"success": False, "error": f"下载视频失败: HTTP {dl_resp.status}"}
|
||||
with open(file_path, "wb") as f:
|
||||
f.write(await dl_resp.read())
|
||||
else:
|
||||
return {"success": False, "error": "Sora 响应中无视频数据"}
|
||||
|
||||
logger.info(f"视频已生成: {file_path} ({os.path.getsize(file_path)} bytes)")
|
||||
return {
|
||||
"success": True,
|
||||
"file_path": file_path,
|
||||
"filename": filename,
|
||||
"url": f"/api/v1/files/{filename}",
|
||||
"model": model,
|
||||
}
|
||||
|
||||
|
||||
# ==================== API 端点 ====================
|
||||
|
||||
@app.get("/")
|
||||
@app.get("/health")
|
||||
async def health_check():
|
||||
return {
|
||||
"status": "healthy",
|
||||
"service": "Ad Creator Agent",
|
||||
"pod_name": POD_NAME,
|
||||
"models": {
|
||||
"image": DEFAULT_IMAGE_MODEL,
|
||||
"text": DEFAULT_TEXT_MODEL,
|
||||
"video": DEFAULT_VIDEO_MODEL,
|
||||
},
|
||||
"callback_enabled": CALLBACK_ENABLED,
|
||||
"timestamp": datetime.utcnow().isoformat(),
|
||||
}
|
||||
|
||||
|
||||
@app.get("/status")
|
||||
async def status():
|
||||
images = list(Path(f"{OUTPUT_DIR}/images").glob("*"))
|
||||
videos = list(Path(f"{OUTPUT_DIR}/videos").glob("*"))
|
||||
return {
|
||||
"status": "running",
|
||||
"pod_name": POD_NAME,
|
||||
"generated_images": len(images),
|
||||
"generated_videos": len(videos),
|
||||
"timestamp": datetime.utcnow().isoformat(),
|
||||
}
|
||||
|
||||
|
||||
# ---------- 图片生成 ----------
|
||||
|
||||
@app.post("/api/v1/generate-image")
|
||||
async def api_generate_image(request: GenerateImageRequest, api_key: str = Depends(get_api_key)):
|
||||
"""生成广告图片"""
|
||||
prompt = request.prompt
|
||||
if request.style:
|
||||
prompt = f"[{request.style} style] {prompt}"
|
||||
if request.brand_name:
|
||||
prompt = f"{prompt}. Brand: {request.brand_name}"
|
||||
|
||||
result = await generate_image_dispatch(
|
||||
prompt=prompt,
|
||||
api_key=api_key,
|
||||
model=request.model,
|
||||
aspect_ratio=request.aspect_ratio or "1:1",
|
||||
size=request.size or "1024x1024",
|
||||
quality=request.quality or "high",
|
||||
reference_image_b64=request.reference_image_b64,
|
||||
)
|
||||
|
||||
if not result.get("success"):
|
||||
raise HTTPException(status_code=500, detail=result.get("error", "图片生成失败"))
|
||||
return result
|
||||
|
||||
|
||||
@app.post("/api/v1/generate-image-upload")
|
||||
async def api_generate_image_with_upload(
|
||||
prompt: str = Form(..., description="广告图片描述"),
|
||||
model: Optional[str] = Form(None),
|
||||
aspect_ratio: Optional[str] = Form("1:1"),
|
||||
quality: Optional[str] = Form("high"),
|
||||
style: Optional[str] = Form(None),
|
||||
brand_name: Optional[str] = Form(None),
|
||||
reference_image: Optional[UploadFile] = File(None, description="参考图片文件"),
|
||||
api_key: str = Depends(get_api_key),
|
||||
):
|
||||
"""上传参考图片并生成广告图（multipart/form-data）"""
|
||||
ref_b64 = None
|
||||
if reference_image:
|
||||
content = await reference_image.read()
|
||||
ref_b64 = base64.b64encode(content).decode("utf-8")
|
||||
|
||||
full_prompt = prompt
|
||||
if style:
|
||||
full_prompt = f"[{style} style] {full_prompt}"
|
||||
if brand_name:
|
||||
full_prompt = f"{full_prompt}. Brand: {brand_name}"
|
||||
|
||||
result = await generate_image_dispatch(
|
||||
prompt=full_prompt,
|
||||
api_key=api_key,
|
||||
model=model,
|
||||
aspect_ratio=aspect_ratio or "1:1",
|
||||
size="1024x1024",
|
||||
quality=quality or "high",
|
||||
reference_image_b64=ref_b64,
|
||||
)
|
||||
|
||||
if not result.get("success"):
|
||||
raise HTTPException(status_code=500, detail=result.get("error", "图片生成失败"))
|
||||
return result
|
||||
|
||||
|
||||
# ---------- 文案生成 ----------
|
||||
|
||||
@app.post("/api/v1/generate-copy")
|
||||
async def api_generate_copy(request: GenerateAdCopyRequest, api_key: str = Depends(get_api_key)):
|
||||
"""生成广告文案"""
|
||||
result = await generate_ad_copy(
|
||||
product=request.product,
|
||||
api_key=api_key,
|
||||
target_audience=request.target_audience,
|
||||
tone=request.tone or "professional",
|
||||
platform=request.platform or "general",
|
||||
language=request.language or "zh",
|
||||
)
|
||||
if not result.get("success"):
|
||||
raise HTTPException(status_code=500, detail=result.get("error", "文案生成失败"))
|
||||
return result
|
||||
|
||||
|
||||
# ---------- 完整广告生成（文案 + 图片） ----------
|
||||
|
||||
@app.post("/api/v1/generate-ad")
|
||||
async def api_generate_ad(request: GenerateAdRequest, api_key: str = Depends(get_api_key)):
|
||||
"""一键生成完整广告（文案 + 配图）"""
|
||||
copy_result = await generate_ad_copy(
|
||||
product=request.product,
|
||||
api_key=api_key,
|
||||
target_audience=request.target_audience,
|
||||
tone=request.tone or "professional",
|
||||
platform=request.platform or "general",
|
||||
language=request.language or "zh",
|
||||
)
|
||||
|
||||
image_prompt = copy_result.get("image_prompt", "")
|
||||
if not image_prompt:
|
||||
image_prompt = f"Advertisement for: {request.product}"
|
||||
|
||||
if request.style:
|
||||
image_prompt = f"[{request.style} style] {image_prompt}"
|
||||
if request.brand_name:
|
||||
image_prompt = f"{image_prompt}. Brand: {request.brand_name}"
|
||||
|
||||
image_result = await generate_image_dispatch(
|
||||
prompt=image_prompt,
|
||||
api_key=api_key,
|
||||
model=request.image_model,
|
||||
aspect_ratio=request.aspect_ratio or "1:1",
|
||||
quality="high",
|
||||
reference_image_b64=request.reference_image_b64,
|
||||
)
|
||||
|
||||
return {
|
||||
"success": True,
|
||||
"copy": copy_result,
|
||||
"image": image_result,
|
||||
"timestamp": datetime.utcnow().isoformat(),
|
||||
}
|
||||
|
||||
|
||||
# ---------- 视频生成 ----------
|
||||
|
||||
@app.post("/api/v1/generate-video")
|
||||
async def api_generate_video(request: GenerateVideoRequest, api_key: str = Depends(get_api_key)):
|
||||
"""生成广告视频"""
|
||||
result = await generate_video_sora(
|
||||
prompt=request.prompt,
|
||||
api_key=api_key,
|
||||
model=request.model,
|
||||
aspect_ratio=request.aspect_ratio or "16:9",
|
||||
duration=request.duration or "5",
|
||||
)
|
||||
if not result.get("success"):
|
||||
raise HTTPException(status_code=500, detail=result.get("error", "视频生成失败"))
|
||||
return result
|
||||
|
||||
|
||||
# ---------- 文件管理 ----------
|
||||
|
||||
@app.get("/api/v1/files/{filename}")
|
||||
async def download_file(filename: str):
|
||||
"""下载生成的文件"""
|
||||
for subdir in ["images", "videos"]:
|
||||
path = os.path.join(OUTPUT_DIR, subdir, filename)
|
||||
if os.path.exists(path):
|
||||
ext = filename.rsplit(".", 1)[-1].lower()
|
||||
media_types = {
|
||||
"png": "image/png", "jpg": "image/jpeg", "jpeg": "image/jpeg",
|
||||
"webp": "image/webp", "mp4": "video/mp4", "gif": "image/gif",
|
||||
}
|
||||
return FileResponse(path, media_type=media_types.get(ext, "application/octet-stream"), filename=filename)
|
||||
raise HTTPException(status_code=404, detail="文件不存在")
|
||||
|
||||
|
||||
@app.get("/api/v1/list-files")
|
||||
async def list_files(file_type: str = "all"):
|
||||
"""列出已生成的文件"""
|
||||
result = {"images": [], "videos": []}
|
||||
|
||||
if file_type in ("all", "image"):
|
||||
img_dir = Path(f"{OUTPUT_DIR}/images")
|
||||
for f in sorted(img_dir.glob("*"), key=lambda p: p.stat().st_mtime, reverse=True):
|
||||
result["images"].append({
|
||||
"filename": f.name,
|
||||
"url": f"/api/v1/files/{f.name}",
|
||||
"size_bytes": f.stat().st_size,
|
||||
"created_at": datetime.fromtimestamp(f.stat().st_mtime).isoformat(),
|
||||
})
|
||||
|
||||
if file_type in ("all", "video"):
|
||||
vid_dir = Path(f"{OUTPUT_DIR}/videos")
|
||||
for f in sorted(vid_dir.glob("*"), key=lambda p: p.stat().st_mtime, reverse=True):
|
||||
result["videos"].append({
|
||||
"filename": f.name,
|
||||
"url": f"/api/v1/files/{f.name}",
|
||||
"size_bytes": f.stat().st_size,
|
||||
"created_at": datetime.fromtimestamp(f.stat().st_mtime).isoformat(),
|
||||
})
|
||||
|
||||
return result
|
||||
|
||||
|
||||
@app.post("/api/v1/cleanup")
|
||||
async def cleanup_files(max_age_hours: int = 24):
|
||||
"""清理超过指定时间的旧文件"""
|
||||
import time
|
||||
|
||||
cutoff = time.time() - max_age_hours * 3600
|
||||
deleted = 0
|
||||
for subdir in ["images", "videos"]:
|
||||
d = Path(f"{OUTPUT_DIR}/{subdir}")
|
||||
for f in d.glob("*"):
|
||||
if f.stat().st_mtime < cutoff:
|
||||
f.unlink()
|
||||
deleted += 1
|
||||
return {"deleted": deleted, "max_age_hours": max_age_hours}
|
||||
|
||||
|
||||
# ---------- 智能对话 ----------
|
||||
|
||||
@app.post("/chat")
|
||||
async def chat(request: ChatRequest, api_key: str = Depends(get_api_key)):
|
||||
"""智能对话 - 理解需求后自动生成广告创意"""
|
||||
|
||||
system_prompt = """你是一位资深广告创意总监 AI 助手。你可以帮用户：
|
||||
1. 分析产品卖点，构思广告创意方案
|
||||
2. 生成广告文案和配图描述
|
||||
3. 推荐合适的广告风格和投放策略
|
||||
|
||||
如果用户想生成图片，在回复末尾添加一行：[GENERATE_IMAGE: 你的英文图片生成 prompt]
|
||||
如果用户只是聊天咨询，正常回复即可。"""
|
||||
|
||||
payload = {
|
||||
"model": DEFAULT_TEXT_MODEL,
|
||||
"messages": [
|
||||
{"role": "system", "content": system_prompt},
|
||||
{"role": "user", "content": request.message},
|
||||
],
|
||||
"max_tokens": 1000,
|
||||
"temperature": 0.8,
|
||||
}
|
||||
|
||||
headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
|
||||
base = LLM_BASE_URL.rstrip("/")
|
||||
url = f"{base}/chat/completions"
|
||||
|
||||
async with aiohttp.ClientSession() as session:
|
||||
async with session.post(url, json=payload, headers=headers, timeout=aiohttp.ClientTimeout(total=30)) as resp:
|
||||
if resp.status != 200:
|
||||
text = await resp.text()
|
||||
raise HTTPException(status_code=500, detail=f"LLM error {resp.status}: {text[:300]}")
|
||||
data = await resp.json()
|
||||
llm_reply = data.get("choices", [{}])[0].get("message", {}).get("content", "")
|
||||
|
||||
image_result = None
|
||||
img_match = re.search(r"\[GENERATE_IMAGE:\s*(.+?)\]", llm_reply)
|
||||
if img_match:
|
||||
img_prompt = img_match.group(1)
|
||||
llm_reply = llm_reply.replace(img_match.group(0), "").strip()
|
||||
image_result = await generate_image_dispatch(prompt=img_prompt, api_key=api_key)
|
||||
|
||||
return {
|
||||
"response": llm_reply,
|
||||
"image": image_result,
|
||||
"timestamp": datetime.utcnow().isoformat(),
|
||||
}
|
||||
|
||||
|
||||
# ==================== 主入口 ====================
|
||||
|
||||
def main():
|
||||
logger.info(f"启动 Ad Creator Agent - {POD_NAME}")
|
||||
logger.info(f"回调功能: {'已启用' if CALLBACK_ENABLED else '未启用'}")
|
||||
uvicorn.run(app, host=SERVICE_HOST, port=SERVICE_PORT, log_level="info")
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -0,0 +1,24 @@
|
||||
__pycache__
|
||||
*.pyc
|
||||
*.pyo
|
||||
*.pyd
|
||||
.Python
|
||||
*.so
|
||||
*.egg
|
||||
*.egg-info
|
||||
dist
|
||||
build
|
||||
.pytest_cache
|
||||
.coverage
|
||||
htmlcov
|
||||
.venv
|
||||
venv
|
||||
outputs/
|
||||
*.mp4
|
||||
*.png
|
||||
*.jpg
|
||||
*.jpeg
|
||||
.DS_Store
|
||||
.env
|
||||
.vscode
|
||||
.idea
|
||||
@@ -0,0 +1,294 @@
|
||||
# Video Generator Agent - 部署和测试指南
|
||||
|
||||
## ✅ 已完成的工作
|
||||
|
||||
### 1. 核心功能实现
|
||||
- ✅ 图片生成工具 (`image_generator.py`)
|
||||
- ✅ 视频处理工具 (`video_processor.py`) - 使用 FFmpeg
|
||||
- ✅ 文件管理工具 (`file_manager.py`)
|
||||
- ✅ MCP 服务器 (`mcp_server.py`) - 4个工具
|
||||
- ✅ API 服务器 (`api_server.py`) - REST API + MCP HTTP/SSE
|
||||
|
||||
### 2. Docker 配置
|
||||
- ✅ Dockerfile（包含 FFmpeg）
|
||||
- ✅ requirements.txt
|
||||
- ✅ .dockerignore
|
||||
- ✅ 健康检查配置
|
||||
|
||||
### 3. 测试和文档
|
||||
- ✅ 测试脚本 (`test_video_agent.py`)
|
||||
- ✅ 核心功能测试 (`test_core_functions.py`)
|
||||
- ✅ README.md 完整文档
|
||||
- ✅ Docker 镜像构建成功
|
||||
- ✅ FFmpeg 验证通过
|
||||
- ✅ 模块导入验证通过
|
||||
|
||||
## 🚀 快速部署
|
||||
|
||||
### 方法 1: Docker 部署（推荐）
|
||||
|
||||
```bash
|
||||
# 1. 构建镜像
|
||||
cd agent_templates/agents/video_generator_agent
|
||||
docker build -t video-generator-agent:latest .
|
||||
|
||||
# 2. 运行容器
|
||||
docker run -d \
|
||||
--name video-generator \
|
||||
-p 8000:8000 \
|
||||
-e OPENAI_API_KEY="sk-i9AwAgXDqqxsA9Ym4AjSPg" \
|
||||
-e MODEL_NAME="dall-e-3" \
|
||||
-v $(pwd)/outputs:/app/outputs \
|
||||
video-generator-agent:latest
|
||||
|
||||
# 3. 查看日志
|
||||
docker logs -f video-generator
|
||||
|
||||
# 4. 测试健康检查
|
||||
curl http://localhost:8000/health
|
||||
```
|
||||
|
||||
### 方法 2: 本地运行
|
||||
|
||||
```bash
|
||||
# 1. 创建虚拟环境
|
||||
python3 -m venv venv
|
||||
source venv/bin/activate
|
||||
|
||||
# 2. 安装依赖
|
||||
pip install -r requirements.txt
|
||||
|
||||
# 3. 安装 FFmpeg（如果未安装）
|
||||
# Ubuntu/Debian:
|
||||
sudo apt-get install ffmpeg
|
||||
|
||||
# macOS:
|
||||
brew install ffmpeg
|
||||
|
||||
# 4. 设置环境变量
|
||||
export OPENAI_API_KEY="sk-i9AwAgXDqqxsA9Ym4AjSPg"
|
||||
export MODEL_NAME="dall-e-3"
|
||||
|
||||
# 5. 启动服务
|
||||
python run_api_server.py
|
||||
```
|
||||
|
||||
## 🧪 测试
|
||||
|
||||
### 1. 健康检查
|
||||
|
||||
```bash
|
||||
curl http://localhost:8000/health
|
||||
```
|
||||
|
||||
预期输出：
|
||||
```json
|
||||
{
|
||||
"status": "healthy",
|
||||
"service": "Video Generator Agent"
|
||||
}
|
||||
```
|
||||
|
||||
### 2. 查看服务信息
|
||||
|
||||
```bash
|
||||
curl http://localhost:8000/
|
||||
```
|
||||
|
||||
### 3. 生成单张图片
|
||||
|
||||
```bash
|
||||
curl -X POST http://localhost:8000/api/v1/generate-image \
|
||||
-H "api-key: sk-i9AwAgXDqqxsA9Ym4AjSPg" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"description": "a beautiful sunset over mountains",
|
||||
"size": "1024x1024"
|
||||
}'
|
||||
```
|
||||
|
||||
### 4. 生成视频
|
||||
|
||||
```bash
|
||||
curl -X POST http://localhost:8000/api/v1/generate-video \
|
||||
-H "api-key: sk-i9AwAgXDqqxsA9Ym4AjSPg" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"descriptions": [
|
||||
"sunrise over mountains",
|
||||
"a peaceful lake at noon",
|
||||
"starry night sky"
|
||||
],
|
||||
"duration_per_image": 3,
|
||||
"transition": "fade"
|
||||
}'
|
||||
```
|
||||
|
||||
### 5. 运行完整测试套件
|
||||
|
||||
```bash
|
||||
# 使用 Docker
|
||||
docker run --rm \
|
||||
-e API_BASE_URL=http://host.docker.internal:8000 \
|
||||
-e API_KEY=sk-i9AwAgXDqqxsA9Ym4AjSPg \
|
||||
--network host \
|
||||
video-generator-agent:latest \
|
||||
python test_video_agent.py
|
||||
|
||||
# 或本地运行
|
||||
python test_video_agent.py
|
||||
```
|
||||
|
||||
## ⚠️ 重要注意事项
|
||||
|
||||
### 1. 模型配置
|
||||
|
||||
**当前问题**: 模型 `taiji/gemini-3-pro-image-preview` 不支持图片生成。
|
||||
|
||||
**解决方案**: 使用以下支持的模型之一：
|
||||
|
||||
```bash
|
||||
# OpenAI DALL-E 模型
|
||||
export MODEL_NAME="dall-e-3"
|
||||
export MODEL_NAME="dall-e-2"
|
||||
|
||||
# 或其他 LiteLLM 支持的图片生成模型
|
||||
```
|
||||
|
||||
### 2. API Key
|
||||
|
||||
确保使用有效的 API Key：
|
||||
```bash
|
||||
export OPENAI_API_KEY="sk-i9AwAgXDqqxsA9Ym4AjSPg"
|
||||
```
|
||||
|
||||
### 3. FFmpeg 依赖
|
||||
|
||||
视频拼接功能需要 FFmpeg。Docker 镜像已包含，本地运行需要手动安装。
|
||||
|
||||
验证 FFmpeg：
|
||||
```bash
|
||||
ffmpeg -version
|
||||
```
|
||||
|
||||
### 4. 存储空间
|
||||
|
||||
生成的图片和视频会占用存储空间。建议：
|
||||
- 定期清理旧文件
|
||||
- 使用 Docker volume 持久化数据
|
||||
- 监控磁盘使用情况
|
||||
|
||||
## 📊 测试结果
|
||||
|
||||
### Docker 构建
|
||||
- ✅ 镜像构建成功
|
||||
- ✅ FFmpeg 7.1.3 已安装
|
||||
- ✅ Python 依赖已安装
|
||||
- ✅ 健康检查配置正确
|
||||
|
||||
### 模块验证
|
||||
- ✅ 所有 Python 模块导入成功
|
||||
- ✅ FFmpeg 可执行
|
||||
- ✅ 服务启动正常
|
||||
|
||||
### 功能状态
|
||||
- ✅ REST API 端点正常
|
||||
- ✅ MCP 端点正常
|
||||
- ✅ 健康检查通过
|
||||
- ⚠️ 图片生成需要正确的模型配置
|
||||
- ✅ 视频拼接功能就绪（FFmpeg）
|
||||
- ✅ 文件管理功能正常
|
||||
|
||||
## 🔧 故障排除
|
||||
|
||||
### 问题 1: 图片生成失败
|
||||
|
||||
**错误**: `not supported model for image generation`
|
||||
|
||||
**解决**:
|
||||
```bash
|
||||
# 更改模型为支持的模型
|
||||
docker run -d \
|
||||
-e MODEL_NAME="dall-e-3" \
|
||||
...
|
||||
```
|
||||
|
||||
### 问题 2: 端口被占用
|
||||
|
||||
**错误**: `Bind for 0.0.0.0:8000 failed: port is already allocated`
|
||||
|
||||
**解决**:
|
||||
```bash
|
||||
# 使用不同端口
|
||||
docker run -d -p 8765:8000 ...
|
||||
```
|
||||
|
||||
### 问题 3: FFmpeg 未找到
|
||||
|
||||
**错误**: `FFmpeg not available`
|
||||
|
||||
**解决**:
|
||||
```bash
|
||||
# Ubuntu/Debian
|
||||
sudo apt-get install ffmpeg
|
||||
|
||||
# macOS
|
||||
brew install ffmpeg
|
||||
```
|
||||
|
||||
## 📝 下一步
|
||||
|
||||
1. **配置正确的图片生成模型**
|
||||
- 联系 LiteLLM Gateway 管理员
|
||||
- 确认可用的图片生成模型
|
||||
- 更新 MODEL_NAME 环境变量
|
||||
|
||||
2. **运行完整测试**
|
||||
```bash
|
||||
python test_video_agent.py
|
||||
```
|
||||
|
||||
3. **集成到 Agent Manager**
|
||||
- 在 `k8s_manager.py` 中添加配置
|
||||
- 在 `app.py` 中注册模板
|
||||
- 部署到 Kubernetes
|
||||
|
||||
4. **生产环境优化**
|
||||
- 配置持久化存储
|
||||
- 设置资源限制
|
||||
- 配置日志收集
|
||||
- 添加监控告警
|
||||
|
||||
## 📚 相关文档
|
||||
|
||||
- [README.md](./README.md) - 完整使用文档
|
||||
- [test_video_agent.py](./test_video_agent.py) - 测试脚本
|
||||
- [Dockerfile](./Dockerfile) - Docker 配置
|
||||
|
||||
## ✨ 功能亮点
|
||||
|
||||
1. **完整的 MCP 支持** - 4个工具，支持 HTTP 和 SSE
|
||||
2. **FFmpeg 视频处理** - 支持多种转场效果
|
||||
3. **文件管理** - 自动清理、存储统计
|
||||
4. **Docker 化** - 开箱即用，包含所有依赖
|
||||
5. **完善的文档** - README、测试脚本、部署指南
|
||||
|
||||
## 🎯 总结
|
||||
|
||||
Video Generator Agent 已经完成开发和基础测试：
|
||||
|
||||
✅ **已完成**:
|
||||
- 核心功能实现（图片生成、视频拼接、文件管理）
|
||||
- MCP 和 REST API 服务器
|
||||
- Docker 配置和构建
|
||||
- 测试脚本和文档
|
||||
- FFmpeg 集成和验证
|
||||
|
||||
⚠️ **待配置**:
|
||||
- 正确的图片生成模型（当前模型不支持）
|
||||
|
||||
🚀 **可以部署**:
|
||||
- Docker 镜像已就绪
|
||||
- 服务可以启动
|
||||
- 视频拼接功能完整
|
||||
- 只需配置正确的模型即可使用图片生成功能
|
||||
@@ -0,0 +1,33 @@
|
||||
FROM python:3.12-slim
|
||||
|
||||
WORKDIR /app
|
||||
|
||||
ENV PYTHONUNBUFFERED=1
|
||||
ENV PYTHONDONTWRITEBYTECODE=1
|
||||
|
||||
# 安装系统依赖（包括 FFmpeg）
|
||||
RUN apt-get update && apt-get install -y \
|
||||
ffmpeg \
|
||||
gcc \
|
||||
curl \
|
||||
&& rm -rf /var/lib/apt/lists/*
|
||||
|
||||
# 验证 FFmpeg 安装
|
||||
RUN ffmpeg -version
|
||||
|
||||
# 安装 Python 依赖
|
||||
COPY requirements.txt .
|
||||
RUN pip install --no-cache-dir -r requirements.txt
|
||||
|
||||
# 复制应用代码
|
||||
COPY . .
|
||||
|
||||
# 创建输出目录
|
||||
RUN mkdir -p /app/outputs/images /app/outputs/videos
|
||||
|
||||
EXPOSE 8000
|
||||
|
||||
HEALTHCHECK --interval=30s --timeout=10s --start-period=5s --retries=3 \
|
||||
CMD curl -f http://localhost:8000/health || exit 1
|
||||
|
||||
CMD ["python", "run_api_server.py"]
|
||||
@@ -0,0 +1,372 @@
|
||||
# Video Generator Agent
|
||||
|
||||
根据文本描述生成图片并拼接为视频的 AI Agent。
|
||||
|
||||
## 功能特性
|
||||
|
||||
- 🎨 **图片生成**: 使用 Gemini 3 Pro Image Preview 根据描述生成高质量图片
|
||||
- 🎬 **视频拼接**: 使用 FFmpeg 将多张图片拼接成视频
|
||||
- ✨ **转场效果**: 支持多种转场效果（fade, wipeleft, wiperight, slideup, slidedown）
|
||||
- 📁 **文件管理**: 自动管理生成的图片和视频文件
|
||||
- 🔌 **MCP 协议**: 支持 MCP (Model Context Protocol) 工具调用
|
||||
- 🌐 **REST API**: 提供完整的 HTTP API 接口
|
||||
|
||||
## 快速开始
|
||||
|
||||
### 1. 本地运行
|
||||
|
||||
```bash
|
||||
# 安装依赖
|
||||
pip install -r requirements.txt
|
||||
|
||||
# 设置环境变量
|
||||
export OPENAI_API_KEY="sk-i9AwAgXDqqxsA9Ym4AjSPg"
|
||||
export MODEL_NAME="taiji/gemini-3-pro-image-preview"
|
||||
|
||||
# 启动服务
|
||||
python run_api_server.py
|
||||
```
|
||||
|
||||
服务将在 `http://localhost:8000` 启动。
|
||||
|
||||
### 2. Docker 运行
|
||||
|
||||
```bash
|
||||
# 构建镜像
|
||||
docker build -t video-generator-agent:latest .
|
||||
|
||||
# 运行容器
|
||||
docker run -d \
|
||||
-p 8000:8000 \
|
||||
-e OPENAI_API_KEY="sk-i9AwAgXDqqxsA9Ym4AjSPg" \
|
||||
-e MODEL_NAME="taiji/gemini-3-pro-image-preview" \
|
||||
-v $(pwd)/outputs:/app/outputs \
|
||||
video-generator-agent:latest
|
||||
```
|
||||
|
||||
### 3. 测试
|
||||
|
||||
```bash
|
||||
# 运行测试脚本
|
||||
python test_video_agent.py
|
||||
|
||||
# 或指定自定义 API URL
|
||||
API_BASE_URL=http://localhost:8000 python test_video_agent.py
|
||||
```
|
||||
|
||||
## API 使用示例
|
||||
|
||||
### 生成单张图片
|
||||
|
||||
```bash
|
||||
curl -X POST http://localhost:8000/api/v1/generate-image \
|
||||
-H "api-key: sk-i9AwAgXDqqxsA9Ym4AjSPg" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"description": "a beautiful sunset over mountains",
|
||||
"size": "1024x1024",
|
||||
"quality": "standard"
|
||||
}'
|
||||
```
|
||||
|
||||
响应示例：
|
||||
```json
|
||||
{
|
||||
"success": true,
|
||||
"file_path": "/app/outputs/images/image_20260226_123456.png",
|
||||
"filename": "image_20260226_123456.png",
|
||||
"url": "/api/v1/files/image_20260226_123456.png",
|
||||
"description": "a beautiful sunset over mountains"
|
||||
}
|
||||
```
|
||||
|
||||
### 生成视频
|
||||
|
||||
```bash
|
||||
curl -X POST http://localhost:8000/api/v1/generate-video \
|
||||
-H "api-key: sk-i9AwAgXDqqxsA9Ym4AjSPg" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"descriptions": [
|
||||
"sunrise over mountains",
|
||||
"a peaceful lake at noon",
|
||||
"starry night sky"
|
||||
],
|
||||
"duration_per_image": 3,
|
||||
"fps": 30,
|
||||
"transition": "fade"
|
||||
}'
|
||||
```
|
||||
|
||||
响应示例：
|
||||
```json
|
||||
{
|
||||
"success": true,
|
||||
"video": {
|
||||
"file_path": "/app/outputs/videos/video_20260226_123456.mp4",
|
||||
"url": "/api/v1/files/video_20260226_123456.mp4",
|
||||
"filename": "video_20260226_123456.mp4",
|
||||
"duration": 9.0,
|
||||
"image_count": 3
|
||||
},
|
||||
"images": [
|
||||
{
|
||||
"index": 1,
|
||||
"description": "sunrise over mountains",
|
||||
"file_path": "/app/outputs/images/image_20260226_123456_1.png",
|
||||
"url": "/api/v1/files/image_20260226_123456_1.png"
|
||||
}
|
||||
],
|
||||
"settings": {
|
||||
"duration_per_image": 3,
|
||||
"fps": 30,
|
||||
"transition": "fade"
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
### 下载文件
|
||||
|
||||
```bash
|
||||
# 下载图片
|
||||
curl -O http://localhost:8000/api/v1/files/image_20260226_123456.png
|
||||
|
||||
# 下载视频
|
||||
curl -O http://localhost:8000/api/v1/files/video_20260226_123456.mp4
|
||||
```
|
||||
|
||||
### 列出文件
|
||||
|
||||
```bash
|
||||
# 列出所有文件
|
||||
curl http://localhost:8000/api/v1/list-files?file_type=all
|
||||
|
||||
# 只列出图片
|
||||
curl http://localhost:8000/api/v1/list-files?file_type=image
|
||||
|
||||
# 只列出视频
|
||||
curl http://localhost:8000/api/v1/list-files?file_type=video
|
||||
```
|
||||
|
||||
## MCP 工具
|
||||
|
||||
Agent 提供以下 MCP 工具：
|
||||
|
||||
### 1. generate_image
|
||||
|
||||
根据描述生成单张图片。
|
||||
|
||||
```json
|
||||
{
|
||||
"name": "generate_image",
|
||||
"arguments": {
|
||||
"description": "a futuristic city at night",
|
||||
"size": "1024x1024",
|
||||
"quality": "standard"
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
### 2. generate_video
|
||||
|
||||
根据多个描述生成视频。
|
||||
|
||||
```json
|
||||
{
|
||||
"name": "generate_video",
|
||||
"arguments": {
|
||||
"descriptions": [
|
||||
"scene 1 description",
|
||||
"scene 2 description",
|
||||
"scene 3 description"
|
||||
],
|
||||
"duration_per_image": 3,
|
||||
"fps": 30,
|
||||
"transition": "fade"
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
### 3. list_generated_files
|
||||
|
||||
列出已生成的文件。
|
||||
|
||||
```json
|
||||
{
|
||||
"name": "list_generated_files",
|
||||
"arguments": {
|
||||
"file_type": "all"
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
### 4. cleanup_old_files
|
||||
|
||||
清理旧文件。
|
||||
|
||||
```json
|
||||
{
|
||||
"name": "cleanup_old_files",
|
||||
"arguments": {
|
||||
"max_age_hours": 24
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
## MCP 端点
|
||||
|
||||
### HTTP 端点
|
||||
|
||||
```bash
|
||||
# 初始化
|
||||
curl -X POST http://localhost:8000/mcp \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"jsonrpc": "2.0",
|
||||
"method": "initialize",
|
||||
"id": 1
|
||||
}'
|
||||
|
||||
# 列出工具
|
||||
curl -X POST http://localhost:8000/mcp \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"jsonrpc": "2.0",
|
||||
"method": "tools/list",
|
||||
"id": 2
|
||||
}'
|
||||
|
||||
# 调用工具
|
||||
curl -X POST http://localhost:8000/mcp \
|
||||
-H "api-key: sk-i9AwAgXDqqxsA9Ym4AjSPg" \
|
||||
-H "Content-Type: application/json" \
|
||||
-d '{
|
||||
"jsonrpc": "2.0",
|
||||
"method": "tools/call",
|
||||
"params": {
|
||||
"name": "generate_image",
|
||||
"arguments": {
|
||||
"description": "a beautiful landscape"
|
||||
}
|
||||
},
|
||||
"id": 3
|
||||
}'
|
||||
```
|
||||
|
||||
### SSE 端点
|
||||
|
||||
```bash
|
||||
# 连接 SSE
|
||||
curl -N http://localhost:8000/mcp/sse
|
||||
```
|
||||
|
||||
## 环境变量
|
||||
|
||||
| 变量 | 必需 | 默认值 | 说明 |
|
||||
|------|------|--------|------|
|
||||
| `OPENAI_API_KEY` | 是 | `sk-i9AwAgXDqqxsA9Ym4AjSPg` | API Key |
|
||||
| `OPENAI_BASE_URL` | 否 | LiteLLM Gateway URL | API Base URL |
|
||||
| `MODEL_NAME` | 否 | `taiji/gemini-3-pro-image-preview` | 模型名称 |
|
||||
| `API_PORT` | 否 | `8000` | 服务端口 |
|
||||
| `OUTPUT_DIR` | 否 | `/app/outputs` | 输出目录 |
|
||||
|
||||
## 转场效果
|
||||
|
||||
支持的转场效果：
|
||||
|
||||
- `fade` - 淡入淡出
|
||||
- `wipeleft` - 左擦除
|
||||
- `wiperight` - 右擦除
|
||||
- `slideup` - 上滑
|
||||
- `slidedown` - 下滑
|
||||
|
||||
## 图片尺寸
|
||||
|
||||
支持的图片尺寸：
|
||||
|
||||
- `256x256`
|
||||
- `512x512`
|
||||
- `1024x1024` (默认)
|
||||
- `1792x1024`
|
||||
- `1024x1792`
|
||||
|
||||
## 项目结构
|
||||
|
||||
```
|
||||
video_generator_agent/
|
||||
├── Dockerfile # Docker 配置
|
||||
├── requirements.txt # Python 依赖
|
||||
├── run_api_server.py # 启动脚本
|
||||
├── test_video_agent.py # 测试脚本
|
||||
├── README.md # 文档
|
||||
└── src/
|
||||
├── __init__.py
|
||||
├── utils/ # 工具模块
|
||||
│ ├── __init__.py
|
||||
│ ├── image_generator.py # 图片生成
|
||||
│ ├── video_processor.py # 视频处理
|
||||
│ └── file_manager.py # 文件管理
|
||||
└── server/ # 服务器模块
|
||||
├── __init__.py
|
||||
├── mcp_server.py # MCP 工具定义
|
||||
└── api_server.py # FastAPI 服务器
|
||||
```
|
||||
|
||||
## 技术栈
|
||||
|
||||
- **Pydantic AI**: Agent 框架
|
||||
- **FastMCP**: MCP 服务器
|
||||
- **FastAPI**: REST API 框架
|
||||
- **FFmpeg**: 视频处理
|
||||
- **Gemini 3 Pro Image Preview**: 图片生成模型
|
||||
- **aiohttp**: 异步 HTTP 客户端
|
||||
- **Pillow**: 图片处理
|
||||
|
||||
## 故障排除
|
||||
|
||||
### FFmpeg 未安装
|
||||
|
||||
如果遇到 FFmpeg 相关错误，请确保已安装 FFmpeg：
|
||||
|
||||
```bash
|
||||
# Ubuntu/Debian
|
||||
sudo apt-get install ffmpeg
|
||||
|
||||
# macOS
|
||||
brew install ffmpeg
|
||||
|
||||
# 验证安装
|
||||
ffmpeg -version
|
||||
```
|
||||
|
||||
### API Key 错误
|
||||
|
||||
确保设置了正确的 API Key：
|
||||
|
||||
```bash
|
||||
export OPENAI_API_KEY="sk-i9AwAgXDqqxsA9Ym4AjSPg"
|
||||
```
|
||||
|
||||
### 端口被占用
|
||||
|
||||
如果端口 8000 被占用，可以更改端口：
|
||||
|
||||
```bash
|
||||
export API_PORT=8080
|
||||
python run_api_server.py
|
||||
```
|
||||
|
||||
## 性能建议
|
||||
|
||||
- 图片生成通常需要 10-30 秒
|
||||
- 视频拼接时间取决于图片数量（每张图片约 2-5 秒）
|
||||
- 建议每个视频不超过 10 个场景
|
||||
- 使用转场效果会增加处理时间
|
||||
|
||||
## 许可证
|
||||
|
||||
MIT License
|
||||
|
||||
## 作者
|
||||
|
||||
Video Generator Agent Team
|
||||
@@ -0,0 +1,19 @@
|
||||
# Pydantic AI
|
||||
pydantic-ai>=0.0.14
|
||||
|
||||
# MCP
|
||||
mcp>=0.9.0
|
||||
fastmcp>=0.1.0
|
||||
|
||||
# FastAPI
|
||||
fastapi>=0.109.0
|
||||
uvicorn[standard]>=0.27.0
|
||||
|
||||
# HTTP Client
|
||||
aiohttp>=3.9.0
|
||||
|
||||
# Image Processing
|
||||
Pillow>=10.0.0
|
||||
|
||||
# OpenAI (for image generation)
|
||||
openai>=1.0.0
|
||||
@@ -0,0 +1,20 @@
|
||||
#!/usr/bin/env python
|
||||
"""启动 Video Generator Agent API 服务器"""
|
||||
import sys
|
||||
from pathlib import Path
|
||||
|
||||
sys.path.insert(0, str(Path(__file__).parent))
|
||||
|
||||
if __name__ == '__main__':
|
||||
from src.server.api_server import app
|
||||
import uvicorn
|
||||
import os
|
||||
|
||||
host = os.getenv('API_HOST', '0.0.0.0')
|
||||
port = int(os.getenv('API_PORT', '8000'))
|
||||
|
||||
print(f"🚀 启动 Video Generator Agent API: http://{host}:{port}")
|
||||
print(f"📖 API 文档: http://{host}:{port}/docs")
|
||||
print(f"🔧 MCP 端点: http://{host}:{port}/mcp")
|
||||
|
||||
uvicorn.run(app, host=host, port=port, log_level="info")
|
||||
@@ -0,0 +1,2 @@
|
||||
"""Video Generator Agent - 视频生成 Agent"""
|
||||
__version__ = "1.0.0"
|
||||
@@ -0,0 +1 @@
|
||||
"""Server modules"""
|
||||
@@ -0,0 +1,345 @@
|
||||
"""
|
||||
HTTP API 服务器 - 视频生成 Agent
|
||||
|
||||
提供 REST API 和 MCP HTTP/SSE 端点。
|
||||
"""
|
||||
import json
|
||||
import uuid
|
||||
import os
|
||||
import sys
|
||||
from typing import Optional, Dict, Any, AsyncGenerator, List
|
||||
from contextlib import asynccontextmanager
|
||||
from pathlib import Path
|
||||
|
||||
# 添加父目录到路径
|
||||
sys.path.insert(0, str(Path(__file__).parent.parent.parent))
|
||||
|
||||
from fastapi import FastAPI, HTTPException, Request, Header, Depends
|
||||
from fastapi.middleware.cors import CORSMiddleware
|
||||
from fastapi.responses import StreamingResponse, JSONResponse, FileResponse
|
||||
from pydantic import BaseModel, Field
|
||||
|
||||
from src.server.mcp_server import TOOL_MAP, TOOL_LIST
|
||||
from src.utils.file_manager import FileManager
|
||||
|
||||
# ==================== 配置 ====================
|
||||
|
||||
SERVER_NAME = "Video Generator Agent"
|
||||
OUTPUT_DIR = os.getenv('OUTPUT_DIR', '/app/outputs')
|
||||
|
||||
file_manager = FileManager(base_dir=OUTPUT_DIR)
|
||||
|
||||
# ==================== FastAPI 应用 ====================
|
||||
|
||||
@asynccontextmanager
|
||||
async def lifespan(app: FastAPI):
|
||||
print(f"🚀 {SERVER_NAME} 启动")
|
||||
print(f"📁 输出目录: {OUTPUT_DIR}")
|
||||
yield
|
||||
print(f"🛑 {SERVER_NAME} 关闭")
|
||||
|
||||
app = FastAPI(
|
||||
title=SERVER_NAME,
|
||||
description="根据描述生成图片并拼接为视频",
|
||||
version="1.0.0",
|
||||
lifespan=lifespan
|
||||
)
|
||||
|
||||
app.add_middleware(
|
||||
CORSMiddleware,
|
||||
allow_origins=["*"],
|
||||
allow_credentials=True,
|
||||
allow_methods=["*"],
|
||||
allow_headers=["*"],
|
||||
)
|
||||
|
||||
|
||||
# ==================== API Key 验证 ====================
|
||||
|
||||
async def verify_api_key(
|
||||
api_key: Optional[str] = Header(None, alias="api-key"),
|
||||
authorization: Optional[str] = Header(None)
|
||||
) -> str:
|
||||
"""验证 API Key"""
|
||||
if api_key and api_key.strip() and api_key.strip() != "sk":
|
||||
return api_key.strip()
|
||||
|
||||
if authorization:
|
||||
key = authorization[7:].strip() if authorization.startswith("Bearer ") else authorization.strip()
|
||||
if key and key != "sk":
|
||||
return key
|
||||
|
||||
raise HTTPException(status_code=401, detail="缺少 API Key")
|
||||
|
||||
|
||||
def get_api_key_from_request(request: Request) -> Optional[str]:
|
||||
"""从请求头提取 API Key（不验证）"""
|
||||
api_key = request.headers.get("api-key") or request.headers.get("api_key")
|
||||
if not api_key:
|
||||
auth = request.headers.get("Authorization")
|
||||
if auth:
|
||||
api_key = auth[7:] if auth.startswith("Bearer ") else auth
|
||||
return api_key
|
||||
|
||||
|
||||
# ==================== 健康检查 ====================
|
||||
|
||||
@app.get("/")
|
||||
async def root():
|
||||
storage_info = file_manager.get_storage_info()
|
||||
return {
|
||||
"service": SERVER_NAME,
|
||||
"status": "running",
|
||||
"version": "1.0.0",
|
||||
"tools": list(TOOL_MAP.keys()),
|
||||
"storage": storage_info
|
||||
}
|
||||
|
||||
|
||||
@app.get("/health")
|
||||
async def health():
|
||||
return {"status": "healthy", "service": SERVER_NAME}
|
||||
|
||||
|
||||
# ==================== MCP 端点 ====================
|
||||
|
||||
sessions: Dict[str, Dict] = {}
|
||||
|
||||
|
||||
async def handle_mcp_request(data: Dict, session_id: str = None, api_key: str = None) -> Dict:
|
||||
"""处理 MCP JSON-RPC 请求"""
|
||||
method = data.get("method")
|
||||
params = data.get("params", {})
|
||||
req_id = data.get("id")
|
||||
|
||||
# tools/call 需要验证 API Key
|
||||
if method == "tools/call" and (not api_key or api_key == "sk"):
|
||||
return {"jsonrpc": "2.0", "id": req_id, "error": {"code": -32001, "message": "缺少 API Key"}}
|
||||
|
||||
try:
|
||||
if method == "initialize":
|
||||
session_id = session_id or str(uuid.uuid4())
|
||||
sessions[session_id] = {"initialized": True}
|
||||
return {
|
||||
"jsonrpc": "2.0", "id": req_id,
|
||||
"result": {
|
||||
"protocolVersion": "2024-11-05",
|
||||
"capabilities": {"tools": {}},
|
||||
"serverInfo": {"name": SERVER_NAME, "version": "1.0.0"}
|
||||
}
|
||||
}
|
||||
|
||||
elif method == "tools/list":
|
||||
return {"jsonrpc": "2.0", "id": req_id, "result": {"tools": TOOL_LIST}}
|
||||
|
||||
elif method == "tools/call":
|
||||
tool_name = params.get("name")
|
||||
args = params.get("arguments", {})
|
||||
|
||||
if tool_name not in TOOL_MAP:
|
||||
raise ValueError(f"Unknown tool: {tool_name}")
|
||||
|
||||
# 设置 API Key 到环境变量
|
||||
old_key = os.environ.get('OPENAI_API_KEY')
|
||||
if api_key:
|
||||
os.environ['OPENAI_API_KEY'] = api_key
|
||||
|
||||
try:
|
||||
result = await TOOL_MAP[tool_name](**args)
|
||||
finally:
|
||||
if old_key:
|
||||
os.environ['OPENAI_API_KEY'] = old_key
|
||||
|
||||
return {
|
||||
"jsonrpc": "2.0", "id": req_id,
|
||||
"result": {"content": [{"type": "text", "text": str(result)}]}
|
||||
}
|
||||
|
||||
elif method == "ping":
|
||||
return {"jsonrpc": "2.0", "id": req_id, "result": {}}
|
||||
|
||||
else:
|
||||
raise ValueError(f"Unknown method: {method}")
|
||||
|
||||
except Exception as e:
|
||||
return {"jsonrpc": "2.0", "id": req_id, "error": {"code": -32603, "message": str(e)}}
|
||||
|
||||
|
||||
@app.post("/mcp")
|
||||
async def mcp_endpoint(request: Request):
|
||||
"""MCP HTTP 端点"""
|
||||
try:
|
||||
body = await request.json()
|
||||
session_id = request.headers.get("x-mcp-session-id")
|
||||
api_key = get_api_key_from_request(request)
|
||||
response = await handle_mcp_request(body, session_id, api_key)
|
||||
return JSONResponse(content=response, headers={"x-mcp-session-id": session_id or ""})
|
||||
except Exception as e:
|
||||
return JSONResponse(status_code=400, content={"jsonrpc": "2.0", "error": {"code": -32700, "message": str(e)}})
|
||||
|
||||
|
||||
@app.get("/mcp/sse")
|
||||
async def mcp_sse(request: Request):
|
||||
"""MCP SSE 端点"""
|
||||
session_id = request.headers.get("x-mcp-session-id") or str(uuid.uuid4())
|
||||
|
||||
async def stream() -> AsyncGenerator[str, None]:
|
||||
yield f"data: {json.dumps({'type': 'connection', 'sessionId': session_id})}\n\n"
|
||||
import asyncio
|
||||
while True:
|
||||
await asyncio.sleep(30)
|
||||
yield f"data: {json.dumps({'type': 'ping'})}\n\n"
|
||||
|
||||
return StreamingResponse(stream(), media_type="text/event-stream",
|
||||
headers={"Cache-Control": "no-cache", "x-mcp-session-id": session_id})
|
||||
|
||||
|
||||
@app.post("/mcp/sse")
|
||||
async def mcp_sse_post(request: Request):
|
||||
"""MCP SSE POST 端点"""
|
||||
try:
|
||||
body = await request.json()
|
||||
session_id = request.headers.get("x-mcp-session-id") or str(uuid.uuid4())
|
||||
api_key = get_api_key_from_request(request)
|
||||
|
||||
async def stream() -> AsyncGenerator[str, None]:
|
||||
response = await handle_mcp_request(body, session_id, api_key)
|
||||
yield f"data: {json.dumps(response)}\n\n"
|
||||
|
||||
return StreamingResponse(stream(), media_type="text/event-stream",
|
||||
headers={"Cache-Control": "no-cache", "x-mcp-session-id": session_id})
|
||||
except Exception as e:
|
||||
return JSONResponse(status_code=400, content={"jsonrpc": "2.0", "error": {"code": -32700, "message": str(e)}})
|
||||
|
||||
|
||||
# ==================== 业务 API ====================
|
||||
|
||||
class ImageGenerationRequest(BaseModel):
|
||||
"""图片生成请求"""
|
||||
description: str = Field(..., description="图片描述")
|
||||
size: str = Field("1024x1024", description="图片尺寸")
|
||||
quality: str = Field("standard", description="图片质量")
|
||||
|
||||
|
||||
class VideoGenerationRequest(BaseModel):
|
||||
"""视频生成请求"""
|
||||
descriptions: List[str] = Field(..., description="场景描述列表")
|
||||
duration_per_image: int = Field(3, description="每张图片显示秒数")
|
||||
fps: int = Field(30, description="视频帧率")
|
||||
transition: Optional[str] = Field(None, description="转场效果")
|
||||
|
||||
|
||||
@app.post("/api/v1/generate-image")
|
||||
async def api_generate_image(request: ImageGenerationRequest, api_key: str = Depends(verify_api_key)):
|
||||
"""生成单张图片"""
|
||||
try:
|
||||
# 设置 API Key
|
||||
old_key = os.environ.get('OPENAI_API_KEY')
|
||||
os.environ['OPENAI_API_KEY'] = api_key
|
||||
|
||||
try:
|
||||
result_str = await TOOL_MAP['generate_image'](
|
||||
description=request.description,
|
||||
size=request.size,
|
||||
quality=request.quality
|
||||
)
|
||||
result = json.loads(result_str)
|
||||
|
||||
if not result.get("success"):
|
||||
raise HTTPException(status_code=500, detail=result.get("error", "Unknown error"))
|
||||
|
||||
return result
|
||||
finally:
|
||||
if old_key:
|
||||
os.environ['OPENAI_API_KEY'] = old_key
|
||||
except HTTPException:
|
||||
raise
|
||||
except Exception as e:
|
||||
raise HTTPException(status_code=500, detail=str(e))
|
||||
|
||||
|
||||
@app.post("/api/v1/generate-video")
|
||||
async def api_generate_video(request: VideoGenerationRequest, api_key: str = Depends(verify_api_key)):
|
||||
"""生成视频"""
|
||||
try:
|
||||
# 设置 API Key
|
||||
old_key = os.environ.get('OPENAI_API_KEY')
|
||||
os.environ['OPENAI_API_KEY'] = api_key
|
||||
|
||||
try:
|
||||
result_str = await TOOL_MAP['generate_video'](
|
||||
descriptions=request.descriptions,
|
||||
duration_per_image=request.duration_per_image,
|
||||
fps=request.fps,
|
||||
transition=request.transition
|
||||
)
|
||||
result = json.loads(result_str)
|
||||
|
||||
if not result.get("success"):
|
||||
raise HTTPException(status_code=500, detail=result.get("error", "Unknown error"))
|
||||
|
||||
return result
|
||||
finally:
|
||||
if old_key:
|
||||
os.environ['OPENAI_API_KEY'] = old_key
|
||||
except HTTPException:
|
||||
raise
|
||||
except Exception as e:
|
||||
raise HTTPException(status_code=500, detail=str(e))
|
||||
|
||||
|
||||
@app.get("/api/v1/list-files")
|
||||
async def api_list_files(file_type: str = "all"):
|
||||
"""列出文件"""
|
||||
try:
|
||||
result_str = await TOOL_MAP['list_generated_files'](file_type=file_type)
|
||||
result = json.loads(result_str)
|
||||
return result
|
||||
except Exception as e:
|
||||
raise HTTPException(status_code=500, detail=str(e))
|
||||
|
||||
|
||||
@app.get("/api/v1/files/{filename}")
|
||||
async def api_download_file(filename: str):
|
||||
"""下载文件"""
|
||||
try:
|
||||
file_path = file_manager.get_file_path(filename)
|
||||
|
||||
if not file_path or not os.path.exists(file_path):
|
||||
raise HTTPException(status_code=404, detail="File not found")
|
||||
|
||||
# 确定 media type
|
||||
if filename.endswith('.mp4'):
|
||||
media_type = "video/mp4"
|
||||
elif filename.endswith('.png'):
|
||||
media_type = "image/png"
|
||||
elif filename.endswith('.jpg') or filename.endswith('.jpeg'):
|
||||
media_type = "image/jpeg"
|
||||
else:
|
||||
media_type = "application/octet-stream"
|
||||
|
||||
return FileResponse(
|
||||
file_path,
|
||||
media_type=media_type,
|
||||
filename=filename
|
||||
)
|
||||
except HTTPException:
|
||||
raise
|
||||
except Exception as e:
|
||||
raise HTTPException(status_code=500, detail=str(e))
|
||||
|
||||
|
||||
@app.post("/api/v1/cleanup")
|
||||
async def api_cleanup(max_age_hours: int = 24):
|
||||
"""清理旧文件"""
|
||||
try:
|
||||
result_str = await TOOL_MAP['cleanup_old_files'](max_age_hours=max_age_hours)
|
||||
result = json.loads(result_str)
|
||||
return result
|
||||
except Exception as e:
|
||||
raise HTTPException(status_code=500, detail=str(e))
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
import uvicorn
|
||||
uvicorn.run(app, host="0.0.0.0", port=8000)
|
||||
@@ -0,0 +1,308 @@
|
||||
"""
|
||||
MCP 服务器 - 视频生成 Agent 工具
|
||||
"""
|
||||
import json
|
||||
import os
|
||||
import sys
|
||||
from typing import Optional, List
|
||||
from pathlib import Path
|
||||
|
||||
# 添加父目录到路径
|
||||
sys.path.insert(0, str(Path(__file__).parent.parent.parent))
|
||||
|
||||
from mcp.server.fastmcp import FastMCP
|
||||
from pydantic_ai import Agent
|
||||
|
||||
from src.utils.image_generator import ImageGenerator
|
||||
from src.utils.video_processor import VideoProcessor
|
||||
from src.utils.file_manager import FileManager
|
||||
|
||||
# ==================== 配置 ====================
|
||||
|
||||
# LiteLLM Gateway 配置
|
||||
_BASE_URL = os.getenv('OPENAI_BASE_URL',
|
||||
os.getenv('LLM_BASE_URL', 'https://litellm.graystone-fb459c5d.southeastasia.azurecontainerapps.io/v1'))
|
||||
_API_KEY = os.getenv('OPENAI_API_KEY', 'sk-i9AwAgXDqqxsA9Ym4AjSPg')
|
||||
|
||||
os.environ.setdefault('OPENAI_API_KEY', _API_KEY)
|
||||
os.environ.setdefault('OPENAI_BASE_URL', _BASE_URL)
|
||||
|
||||
# 模型名称
|
||||
def _get_model_name() -> str:
|
||||
model = os.getenv('MODEL_NAME', os.getenv('LITELLM_MODEL', 'taiji/gemini-3-pro-image-preview'))
|
||||
return model if ':' in model else f'openai:{model}'
|
||||
|
||||
MODEL_NAME = _get_model_name()
|
||||
|
||||
# 输出目录
|
||||
OUTPUT_DIR = os.getenv('OUTPUT_DIR', '/app/outputs')
|
||||
|
||||
# ==================== 初始化工具 ====================
|
||||
|
||||
image_generator = ImageGenerator(api_key=_API_KEY, base_url=_BASE_URL, model='taiji/gemini-3-pro-image-preview')
|
||||
video_processor = VideoProcessor()
|
||||
file_manager = FileManager(base_dir=OUTPUT_DIR)
|
||||
|
||||
# ==================== MCP 服务器 ====================
|
||||
|
||||
server = FastMCP('Video Generator Agent')
|
||||
|
||||
SYSTEM_PROMPT = '''你是一个专业的视频生成 AI 助手。
|
||||
你可以根据用户的描述生成图片，并将多张图片拼接成视频。
|
||||
请根据用户的需求提供帮助。'''
|
||||
|
||||
|
||||
def get_agent() -> Agent:
|
||||
"""创建 Agent 实例"""
|
||||
return Agent(MODEL_NAME, system_prompt=SYSTEM_PROMPT)
|
||||
|
||||
|
||||
# ==================== MCP 工具定义 ====================
|
||||
|
||||
@server.tool()
|
||||
async def generate_image(
|
||||
description: str,
|
||||
size: str = "1024x1024",
|
||||
quality: str = "standard"
|
||||
) -> str:
|
||||
"""
|
||||
根据描述生成单张图片
|
||||
|
||||
Args:
|
||||
description: 图片描述（英文效果更好）
|
||||
size: 图片尺寸，可选 256x256, 512x512, 1024x1024, 1792x1024, 1024x1792
|
||||
quality: 图片质量，可选 standard, hd
|
||||
|
||||
Returns:
|
||||
JSON 格式的结果，包含图片路径和下载 URL
|
||||
"""
|
||||
try:
|
||||
result = await image_generator.generate_image(
|
||||
description=description,
|
||||
output_dir=os.path.join(OUTPUT_DIR, "images"),
|
||||
size=size,
|
||||
quality=quality
|
||||
)
|
||||
|
||||
return json.dumps(result, ensure_ascii=False, indent=2)
|
||||
|
||||
except Exception as e:
|
||||
return json.dumps({
|
||||
"success": False,
|
||||
"error": str(e)
|
||||
}, ensure_ascii=False)
|
||||
|
||||
|
||||
@server.tool()
|
||||
async def generate_video(
|
||||
descriptions: List[str],
|
||||
duration_per_image: int = 3,
|
||||
fps: int = 30,
|
||||
transition: Optional[str] = None
|
||||
) -> str:
|
||||
"""
|
||||
根据多个描述生成视频
|
||||
|
||||
Args:
|
||||
descriptions: 场景描述列表（每个描述对应一张图片）
|
||||
duration_per_image: 每张图片显示秒数（默认 3 秒）
|
||||
fps: 视频帧率（默认 30）
|
||||
transition: 转场效果，可选 fade, wipeleft, wiperight, slideup, slidedown
|
||||
|
||||
Returns:
|
||||
JSON 格式的结果，包含视频路径和下载 URL
|
||||
"""
|
||||
try:
|
||||
if not descriptions or len(descriptions) == 0:
|
||||
return json.dumps({
|
||||
"success": False,
|
||||
"error": "至少需要提供一个场景描述"
|
||||
}, ensure_ascii=False)
|
||||
|
||||
# 1. 生成所有图片
|
||||
image_paths = []
|
||||
generated_images = []
|
||||
|
||||
for i, desc in enumerate(descriptions):
|
||||
print(f"生成图片 {i+1}/{len(descriptions)}: {desc[:50]}...")
|
||||
|
||||
result = await image_generator.generate_image(
|
||||
description=desc,
|
||||
output_dir=os.path.join(OUTPUT_DIR, "images")
|
||||
)
|
||||
|
||||
if not result.get("success"):
|
||||
return json.dumps({
|
||||
"success": False,
|
||||
"error": f"生成第 {i+1} 张图片失败: {result.get('error')}",
|
||||
"generated_images": generated_images
|
||||
}, ensure_ascii=False)
|
||||
|
||||
image_paths.append(result["file_path"])
|
||||
generated_images.append({
|
||||
"index": i + 1,
|
||||
"description": desc,
|
||||
"file_path": result["file_path"],
|
||||
"url": result["url"]
|
||||
})
|
||||
|
||||
# 2. 拼接视频
|
||||
print(f"拼接视频: {len(image_paths)} 张图片...")
|
||||
|
||||
video_result = await video_processor.create_video_from_images(
|
||||
image_paths=image_paths,
|
||||
output_dir=os.path.join(OUTPUT_DIR, "videos"),
|
||||
duration_per_image=duration_per_image,
|
||||
fps=fps,
|
||||
transition=transition
|
||||
)
|
||||
|
||||
if not video_result.get("success"):
|
||||
return json.dumps({
|
||||
"success": False,
|
||||
"error": f"视频拼接失败: {video_result.get('error')}",
|
||||
"generated_images": generated_images
|
||||
}, ensure_ascii=False)
|
||||
|
||||
# 3. 返回结果
|
||||
return json.dumps({
|
||||
"success": True,
|
||||
"video": {
|
||||
"file_path": video_result["file_path"],
|
||||
"url": video_result["url"],
|
||||
"filename": video_result["filename"],
|
||||
"duration": video_result["duration"],
|
||||
"image_count": video_result["image_count"]
|
||||
},
|
||||
"images": generated_images,
|
||||
"settings": {
|
||||
"duration_per_image": duration_per_image,
|
||||
"fps": fps,
|
||||
"transition": transition
|
||||
}
|
||||
}, ensure_ascii=False, indent=2)
|
||||
|
||||
except Exception as e:
|
||||
return json.dumps({
|
||||
"success": False,
|
||||
"error": str(e)
|
||||
}, ensure_ascii=False)
|
||||
|
||||
|
||||
@server.tool()
|
||||
async def list_generated_files(file_type: str = "all") -> str:
|
||||
"""
|
||||
列出已生成的文件
|
||||
|
||||
Args:
|
||||
file_type: 文件类型，可选 all, image, video
|
||||
|
||||
Returns:
|
||||
JSON 格式的文件列表
|
||||
"""
|
||||
try:
|
||||
files = file_manager.list_files(file_type=file_type)
|
||||
storage_info = file_manager.get_storage_info()
|
||||
|
||||
return json.dumps({
|
||||
"success": True,
|
||||
"files": files,
|
||||
"storage": storage_info,
|
||||
"total_count": len(files)
|
||||
}, ensure_ascii=False, indent=2)
|
||||
|
||||
except Exception as e:
|
||||
return json.dumps({
|
||||
"success": False,
|
||||
"error": str(e)
|
||||
}, ensure_ascii=False)
|
||||
|
||||
|
||||
@server.tool()
|
||||
async def cleanup_old_files(max_age_hours: int = 24) -> str:
|
||||
"""
|
||||
清理旧文件
|
||||
|
||||
Args:
|
||||
max_age_hours: 保留最近多少小时的文件（默认 24 小时）
|
||||
|
||||
Returns:
|
||||
JSON 格式的清理结果
|
||||
"""
|
||||
try:
|
||||
result = file_manager.cleanup_old_files(max_age_hours=max_age_hours)
|
||||
return json.dumps(result, ensure_ascii=False, indent=2)
|
||||
|
||||
except Exception as e:
|
||||
return json.dumps({
|
||||
"success": False,
|
||||
"error": str(e)
|
||||
}, ensure_ascii=False)
|
||||
|
||||
|
||||
# ==================== 工具映射（供 API 使用）====================
|
||||
|
||||
TOOL_MAP = {
|
||||
'generate_image': generate_image,
|
||||
'generate_video': generate_video,
|
||||
'list_generated_files': list_generated_files,
|
||||
'cleanup_old_files': cleanup_old_files,
|
||||
}
|
||||
|
||||
TOOL_LIST = [
|
||||
{
|
||||
"name": "generate_image",
|
||||
"description": "根据描述生成单张图片",
|
||||
"inputSchema": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"description": {"type": "string", "description": "图片描述"},
|
||||
"size": {"type": "string", "description": "图片尺寸", "default": "1024x1024"},
|
||||
"quality": {"type": "string", "description": "图片质量", "default": "standard"}
|
||||
},
|
||||
"required": ["description"]
|
||||
}
|
||||
},
|
||||
{
|
||||
"name": "generate_video",
|
||||
"description": "根据多个描述生成视频",
|
||||
"inputSchema": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"descriptions": {
|
||||
"type": "array",
|
||||
"items": {"type": "string"},
|
||||
"description": "场景描述列表"
|
||||
},
|
||||
"duration_per_image": {"type": "integer", "description": "每张图片显示秒数", "default": 3},
|
||||
"fps": {"type": "integer", "description": "视频帧率", "default": 30},
|
||||
"transition": {"type": "string", "description": "转场效果"}
|
||||
},
|
||||
"required": ["descriptions"]
|
||||
}
|
||||
},
|
||||
{
|
||||
"name": "list_generated_files",
|
||||
"description": "列出已生成的文件",
|
||||
"inputSchema": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"file_type": {"type": "string", "description": "文件类型 (all/image/video)", "default": "all"}
|
||||
}
|
||||
}
|
||||
},
|
||||
{
|
||||
"name": "cleanup_old_files",
|
||||
"description": "清理旧文件",
|
||||
"inputSchema": {
|
||||
"type": "object",
|
||||
"properties": {
|
||||
"max_age_hours": {"type": "integer", "description": "保留最近多少小时的文件", "default": 24}
|
||||
}
|
||||
}
|
||||
}
|
||||
]
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
server.run()
|
||||
@@ -0,0 +1,6 @@
|
||||
"""Utility modules for video generation"""
|
||||
from .image_generator import ImageGenerator
|
||||
from .video_processor import VideoProcessor
|
||||
from .file_manager import FileManager
|
||||
|
||||
__all__ = ['ImageGenerator', 'VideoProcessor', 'FileManager']
|
||||
@@ -0,0 +1,218 @@
|
||||
"""
|
||||
文件管理工具
|
||||
管理生成的图片和视频文件
|
||||
"""
|
||||
import os
|
||||
import logging
|
||||
from pathlib import Path
|
||||
from typing import List, Dict, Optional
|
||||
from datetime import datetime, timedelta
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
class FileManager:
|
||||
"""文件管理器"""
|
||||
|
||||
def __init__(self, base_dir: str = "/app/outputs"):
|
||||
"""
|
||||
初始化文件管理器
|
||||
|
||||
Args:
|
||||
base_dir: 基础目录
|
||||
"""
|
||||
self.base_dir = base_dir
|
||||
self.images_dir = os.path.join(base_dir, "images")
|
||||
self.videos_dir = os.path.join(base_dir, "videos")
|
||||
|
||||
# 创建目录
|
||||
Path(self.images_dir).mkdir(parents=True, exist_ok=True)
|
||||
Path(self.videos_dir).mkdir(parents=True, exist_ok=True)
|
||||
|
||||
logger.info(f"FileManager initialized: base_dir={base_dir}")
|
||||
|
||||
def get_file_path(self, filename: str) -> Optional[str]:
|
||||
"""
|
||||
获取文件完整路径
|
||||
|
||||
Args:
|
||||
filename: 文件名
|
||||
|
||||
Returns:
|
||||
str: 文件路径，如果不存在返回 None
|
||||
"""
|
||||
# 先在 images 目录查找
|
||||
image_path = os.path.join(self.images_dir, filename)
|
||||
if os.path.exists(image_path):
|
||||
return image_path
|
||||
|
||||
# 再在 videos 目录查找
|
||||
video_path = os.path.join(self.videos_dir, filename)
|
||||
if os.path.exists(video_path):
|
||||
return video_path
|
||||
|
||||
return None
|
||||
|
||||
def list_files(self, file_type: str = "all") -> List[Dict]:
|
||||
"""
|
||||
列出文件
|
||||
|
||||
Args:
|
||||
file_type: 文件类型 (all, image, video)
|
||||
|
||||
Returns:
|
||||
List[Dict]: 文件列表
|
||||
"""
|
||||
files = []
|
||||
|
||||
try:
|
||||
if file_type in ["all", "image"]:
|
||||
files.extend(self._list_directory(self.images_dir, "image"))
|
||||
|
||||
if file_type in ["all", "video"]:
|
||||
files.extend(self._list_directory(self.videos_dir, "video"))
|
||||
|
||||
# 按修改时间排序（最新的在前）
|
||||
files.sort(key=lambda x: x["modified_time"], reverse=True)
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error listing files: {e}")
|
||||
|
||||
return files
|
||||
|
||||
def _list_directory(self, directory: str, file_type: str) -> List[Dict]:
|
||||
"""
|
||||
列出目录中的文件
|
||||
|
||||
Args:
|
||||
directory: 目录路径
|
||||
file_type: 文件类型标签
|
||||
|
||||
Returns:
|
||||
List[Dict]: 文件信息列表
|
||||
"""
|
||||
files = []
|
||||
|
||||
try:
|
||||
if not os.path.exists(directory):
|
||||
return files
|
||||
|
||||
for filename in os.listdir(directory):
|
||||
file_path = os.path.join(directory, filename)
|
||||
|
||||
if not os.path.isfile(file_path):
|
||||
continue
|
||||
|
||||
stat = os.stat(file_path)
|
||||
|
||||
files.append({
|
||||
"filename": filename,
|
||||
"type": file_type,
|
||||
"size": stat.st_size,
|
||||
"size_mb": round(stat.st_size / 1024 / 1024, 2),
|
||||
"modified_time": stat.st_mtime,
|
||||
"modified_date": datetime.fromtimestamp(stat.st_mtime).isoformat(),
|
||||
"url": f"/api/v1/files/{filename}"
|
||||
})
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error listing directory {directory}: {e}")
|
||||
|
||||
return files
|
||||
|
||||
def cleanup_old_files(self, max_age_hours: int = 24) -> Dict:
|
||||
"""
|
||||
清理旧文件
|
||||
|
||||
Args:
|
||||
max_age_hours: 最大保留时间（小时）
|
||||
|
||||
Returns:
|
||||
Dict: 清理结果
|
||||
"""
|
||||
try:
|
||||
cutoff_time = datetime.now() - timedelta(hours=max_age_hours)
|
||||
cutoff_timestamp = cutoff_time.timestamp()
|
||||
|
||||
deleted_count = 0
|
||||
freed_space = 0
|
||||
|
||||
# 清理图片
|
||||
for filename in os.listdir(self.images_dir):
|
||||
file_path = os.path.join(self.images_dir, filename)
|
||||
if os.path.isfile(file_path):
|
||||
stat = os.stat(file_path)
|
||||
if stat.st_mtime < cutoff_timestamp:
|
||||
freed_space += stat.st_size
|
||||
os.remove(file_path)
|
||||
deleted_count += 1
|
||||
logger.info(f"Deleted old image: {filename}")
|
||||
|
||||
# 清理视频
|
||||
for filename in os.listdir(self.videos_dir):
|
||||
file_path = os.path.join(self.videos_dir, filename)
|
||||
if os.path.isfile(file_path):
|
||||
stat = os.stat(file_path)
|
||||
if stat.st_mtime < cutoff_timestamp:
|
||||
freed_space += stat.st_size
|
||||
os.remove(file_path)
|
||||
deleted_count += 1
|
||||
logger.info(f"Deleted old video: {filename}")
|
||||
|
||||
return {
|
||||
"success": True,
|
||||
"deleted_count": deleted_count,
|
||||
"freed_space_mb": round(freed_space / 1024 / 1024, 2)
|
||||
}
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error cleaning up files: {e}")
|
||||
return {
|
||||
"success": False,
|
||||
"error": str(e)
|
||||
}
|
||||
|
||||
def get_storage_info(self) -> Dict:
|
||||
"""
|
||||
获取存储信息
|
||||
|
||||
Returns:
|
||||
Dict: 存储信息
|
||||
"""
|
||||
try:
|
||||
total_size = 0
|
||||
image_count = 0
|
||||
video_count = 0
|
||||
|
||||
# 统计图片
|
||||
if os.path.exists(self.images_dir):
|
||||
for filename in os.listdir(self.images_dir):
|
||||
file_path = os.path.join(self.images_dir, filename)
|
||||
if os.path.isfile(file_path):
|
||||
total_size += os.path.getsize(file_path)
|
||||
image_count += 1
|
||||
|
||||
# 统计视频
|
||||
if os.path.exists(self.videos_dir):
|
||||
for filename in os.listdir(self.videos_dir):
|
||||
file_path = os.path.join(self.videos_dir, filename)
|
||||
if os.path.isfile(file_path):
|
||||
total_size += os.path.getsize(file_path)
|
||||
video_count += 1
|
||||
|
||||
return {
|
||||
"total_size_mb": round(total_size / 1024 / 1024, 2),
|
||||
"image_count": image_count,
|
||||
"video_count": video_count,
|
||||
"total_files": image_count + video_count
|
||||
}
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error getting storage info: {e}")
|
||||
return {
|
||||
"total_size_mb": 0,
|
||||
"image_count": 0,
|
||||
"video_count": 0,
|
||||
"total_files": 0,
|
||||
"error": str(e)
|
||||
}
|
||||
@@ -0,0 +1,197 @@
|
||||
"""
|
||||
图片生成工具
|
||||
使用 OpenAI 兼容接口调用 Gemini 3 Pro Image Preview
|
||||
"""
|
||||
import os
|
||||
import aiohttp
|
||||
import logging
|
||||
from pathlib import Path
|
||||
from typing import Optional
|
||||
from datetime import datetime
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
class ImageGenerator:
|
||||
"""图片生成器"""
|
||||
|
||||
def __init__(self, api_key: str = None, base_url: str = None, model: str = None):
|
||||
"""
|
||||
初始化图片生成器
|
||||
|
||||
Args:
|
||||
api_key: API Key
|
||||
base_url: API Base URL
|
||||
model: 模型名称
|
||||
"""
|
||||
self.api_key = api_key or os.getenv('OPENAI_API_KEY', 'sk')
|
||||
self.base_url = base_url or os.getenv('OPENAI_BASE_URL',
|
||||
'https://litellm.graystone-fb459c5d.southeastasia.azurecontainerapps.io/v1')
|
||||
self.model = model or os.getenv('MODEL_NAME', 'taiji/gemini-3-pro-image-preview')
|
||||
|
||||
# 确保 base_url 不以 /v1 结尾（我们会手动添加）
|
||||
if self.base_url.endswith('/v1'):
|
||||
self.base_url = self.base_url[:-3]
|
||||
|
||||
logger.info(f"ImageGenerator initialized: model={self.model}, base_url={self.base_url}")
|
||||
|
||||
async def generate_image(
|
||||
self,
|
||||
description: str,
|
||||
output_dir: str = "/app/outputs/images",
|
||||
size: str = "1024x1024",
|
||||
quality: str = "standard"
|
||||
) -> dict:
|
||||
"""
|
||||
生成图片
|
||||
|
||||
Args:
|
||||
description: 图片描述
|
||||
output_dir: 输出目录
|
||||
size: 图片尺寸
|
||||
quality: 图片质量
|
||||
|
||||
Returns:
|
||||
dict: {
|
||||
"success": bool,
|
||||
"file_path": str,
|
||||
"url": str,
|
||||
"description": str
|
||||
}
|
||||
"""
|
||||
try:
|
||||
# 创建输出目录
|
||||
Path(output_dir).mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# 调用 API 生成图片
|
||||
logger.info(f"Generating image: {description[:50]}...")
|
||||
|
||||
async with aiohttp.ClientSession() as session:
|
||||
# 使用 OpenAI 兼容的图片生成接口
|
||||
url = f"{self.base_url}/v1/images/generations"
|
||||
headers = {
|
||||
"Authorization": f"Bearer {self.api_key}",
|
||||
"Content-Type": "application/json"
|
||||
}
|
||||
payload = {
|
||||
"model": self.model,
|
||||
"prompt": description,
|
||||
"n": 1,
|
||||
"size": size,
|
||||
"quality": quality
|
||||
}
|
||||
|
||||
async with session.post(url, json=payload, headers=headers, timeout=aiohttp.ClientTimeout(total=60)) as response:
|
||||
if response.status != 200:
|
||||
error_text = await response.text()
|
||||
logger.error(f"API error: {response.status} - {error_text}")
|
||||
return {
|
||||
"success": False,
|
||||
"error": f"API error: {response.status} - {error_text}"
|
||||
}
|
||||
|
||||
data = await response.json()
|
||||
|
||||
# 获取图片 URL
|
||||
if not data.get("data") or len(data["data"]) == 0:
|
||||
return {
|
||||
"success": False,
|
||||
"error": "No image data returned from API"
|
||||
}
|
||||
|
||||
image_url = data["data"][0].get("url")
|
||||
if not image_url:
|
||||
return {
|
||||
"success": False,
|
||||
"error": "No image URL in response"
|
||||
}
|
||||
|
||||
# 下载图片
|
||||
timestamp = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
|
||||
filename = f"image_{timestamp}.png"
|
||||
file_path = os.path.join(output_dir, filename)
|
||||
|
||||
download_result = await self._download_image(image_url, file_path)
|
||||
if not download_result["success"]:
|
||||
return download_result
|
||||
|
||||
logger.info(f"Image generated successfully: {file_path}")
|
||||
|
||||
return {
|
||||
"success": True,
|
||||
"file_path": file_path,
|
||||
"filename": filename,
|
||||
"url": f"/api/v1/files/{filename}",
|
||||
"description": description
|
||||
}
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error generating image: {e}", exc_info=True)
|
||||
return {
|
||||
"success": False,
|
||||
"error": str(e)
|
||||
}
|
||||
|
||||
async def _download_image(self, url: str, save_path: str) -> dict:
|
||||
"""
|
||||
下载图片到本地
|
||||
|
||||
Args:
|
||||
url: 图片 URL
|
||||
save_path: 保存路径
|
||||
|
||||
Returns:
|
||||
dict: {"success": bool, "file_path": str, "error": str}
|
||||
"""
|
||||
try:
|
||||
async with aiohttp.ClientSession() as session:
|
||||
async with session.get(url, timeout=aiohttp.ClientTimeout(total=30)) as response:
|
||||
if response.status != 200:
|
||||
return {
|
||||
"success": False,
|
||||
"error": f"Failed to download image: HTTP {response.status}"
|
||||
}
|
||||
|
||||
content = await response.read()
|
||||
|
||||
# 保存文件
|
||||
with open(save_path, 'wb') as f:
|
||||
f.write(content)
|
||||
|
||||
# 验证文件
|
||||
if not os.path.exists(save_path) or os.path.getsize(save_path) == 0:
|
||||
return {
|
||||
"success": False,
|
||||
"error": "Downloaded file is empty or not saved"
|
||||
}
|
||||
|
||||
return {
|
||||
"success": True,
|
||||
"file_path": save_path
|
||||
}
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error downloading image: {e}")
|
||||
return {
|
||||
"success": False,
|
||||
"error": str(e)
|
||||
}
|
||||
|
||||
def validate_image(self, file_path: str) -> bool:
|
||||
"""
|
||||
验证图片文件是否有效
|
||||
|
||||
Args:
|
||||
file_path: 文件路径
|
||||
|
||||
Returns:
|
||||
bool: 是否有效
|
||||
"""
|
||||
try:
|
||||
from PIL import Image
|
||||
with Image.open(file_path) as img:
|
||||
img.verify()
|
||||
return True
|
||||
except Exception as e:
|
||||
logger.error(f"Image validation failed: {e}")
|
||||
return False
|
||||
@@ -0,0 +1,327 @@
|
||||
"""
|
||||
视频处理工具
|
||||
使用 FFmpeg 将图片拼接成视频
|
||||
"""
|
||||
import os
|
||||
import subprocess
|
||||
import logging
|
||||
from pathlib import Path
|
||||
from typing import List, Optional
|
||||
from datetime import datetime
|
||||
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
|
||||
class VideoProcessor:
|
||||
"""视频处理器"""
|
||||
|
||||
def __init__(self):
|
||||
"""初始化视频处理器"""
|
||||
# 检查 FFmpeg 是否可用
|
||||
try:
|
||||
result = subprocess.run(['ffmpeg', '-version'],
|
||||
capture_output=True, text=True, timeout=5)
|
||||
if result.returncode == 0:
|
||||
logger.info("FFmpeg is available")
|
||||
else:
|
||||
logger.warning("FFmpeg check returned non-zero exit code")
|
||||
except Exception as e:
|
||||
logger.error(f"FFmpeg not available: {e}")
|
||||
|
||||
async def create_video_from_images(
|
||||
self,
|
||||
image_paths: List[str],
|
||||
output_dir: str = "/app/outputs/videos",
|
||||
duration_per_image: int = 3,
|
||||
fps: int = 30,
|
||||
transition: Optional[str] = None
|
||||
) -> dict:
|
||||
"""
|
||||
从图片创建视频
|
||||
|
||||
Args:
|
||||
image_paths: 图片路径列表
|
||||
output_dir: 输出目录
|
||||
duration_per_image: 每张图片显示秒数
|
||||
fps: 帧率
|
||||
transition: 转场效果 (fade, wipeleft, wiperight, slideup, slidedown)
|
||||
|
||||
Returns:
|
||||
dict: {
|
||||
"success": bool,
|
||||
"file_path": str,
|
||||
"url": str,
|
||||
"duration": float,
|
||||
"image_count": int
|
||||
}
|
||||
"""
|
||||
try:
|
||||
# 验证输入
|
||||
if not image_paths or len(image_paths) == 0:
|
||||
return {
|
||||
"success": False,
|
||||
"error": "No images provided"
|
||||
}
|
||||
|
||||
# 验证所有图片文件存在
|
||||
for img_path in image_paths:
|
||||
if not os.path.exists(img_path):
|
||||
return {
|
||||
"success": False,
|
||||
"error": f"Image not found: {img_path}"
|
||||
}
|
||||
|
||||
# 创建输出目录
|
||||
Path(output_dir).mkdir(parents=True, exist_ok=True)
|
||||
|
||||
# 生成输出文件名
|
||||
timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
|
||||
output_filename = f"video_{timestamp}.mp4"
|
||||
output_path = os.path.join(output_dir, output_filename)
|
||||
|
||||
logger.info(f"Creating video from {len(image_paths)} images...")
|
||||
|
||||
# 根据是否需要转场效果选择不同的方法
|
||||
if transition and len(image_paths) > 1:
|
||||
result = await self._create_video_with_transitions(
|
||||
image_paths, output_path, duration_per_image, fps, transition
|
||||
)
|
||||
else:
|
||||
result = await self._create_simple_video(
|
||||
image_paths, output_path, duration_per_image, fps
|
||||
)
|
||||
|
||||
if not result["success"]:
|
||||
return result
|
||||
|
||||
# 获取视频信息
|
||||
video_info = self.get_video_info(output_path)
|
||||
|
||||
return {
|
||||
"success": True,
|
||||
"file_path": output_path,
|
||||
"filename": output_filename,
|
||||
"url": f"/api/v1/files/{output_filename}",
|
||||
"duration": video_info.get("duration", 0),
|
||||
"image_count": len(image_paths)
|
||||
}
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error creating video: {e}", exc_info=True)
|
||||
return {
|
||||
"success": False,
|
||||
"error": str(e)
|
||||
}
|
||||
|
||||
async def _create_simple_video(
|
||||
self,
|
||||
image_paths: List[str],
|
||||
output_path: str,
|
||||
duration_per_image: int,
|
||||
fps: int
|
||||
) -> dict:
|
||||
"""
|
||||
创建简单视频（无转场效果）
|
||||
|
||||
使用 concat demuxer 方法
|
||||
"""
|
||||
try:
|
||||
# 创建临时文件列表
|
||||
temp_dir = os.path.dirname(output_path)
|
||||
concat_file = os.path.join(temp_dir, f"concat_{datetime.now().strftime('%Y%m%d_%H%M%S')}.txt")
|
||||
|
||||
# 写入文件列表
|
||||
with open(concat_file, 'w') as f:
|
||||
for img_path in image_paths:
|
||||
# FFmpeg concat 格式
|
||||
f.write(f"file '{img_path}'\n")
|
||||
f.write(f"duration {duration_per_image}\n")
|
||||
# 最后一张图片需要再写一次（FFmpeg 要求）
|
||||
f.write(f"file '{image_paths[-1]}'\n")
|
||||
|
||||
# 构建 FFmpeg 命令
|
||||
cmd = [
|
||||
'ffmpeg',
|
||||
'-f', 'concat',
|
||||
'-safe', '0',
|
||||
'-i', concat_file,
|
||||
'-vsync', 'vfr',
|
||||
'-pix_fmt', 'yuv420p',
|
||||
'-c:v', 'libx264',
|
||||
'-r', str(fps),
|
||||
'-y', # 覆盖输出文件
|
||||
output_path
|
||||
]
|
||||
|
||||
logger.info(f"Running FFmpeg command: {' '.join(cmd)}")
|
||||
|
||||
# 执行命令
|
||||
result = subprocess.run(
|
||||
cmd,
|
||||
capture_output=True,
|
||||
text=True,
|
||||
timeout=300 # 5分钟超时
|
||||
)
|
||||
|
||||
# 清理临时文件
|
||||
try:
|
||||
os.remove(concat_file)
|
||||
except:
|
||||
pass
|
||||
|
||||
if result.returncode != 0:
|
||||
logger.error(f"FFmpeg error: {result.stderr}")
|
||||
return {
|
||||
"success": False,
|
||||
"error": f"FFmpeg failed: {result.stderr[:500]}"
|
||||
}
|
||||
|
||||
# 验证输出文件
|
||||
if not os.path.exists(output_path) or os.path.getsize(output_path) == 0:
|
||||
return {
|
||||
"success": False,
|
||||
"error": "Output video file is empty or not created"
|
||||
}
|
||||
|
||||
logger.info(f"Video created successfully: {output_path}")
|
||||
return {"success": True}
|
||||
|
||||
except subprocess.TimeoutExpired:
|
||||
return {
|
||||
"success": False,
|
||||
"error": "FFmpeg command timed out"
|
||||
}
|
||||
except Exception as e:
|
||||
logger.error(f"Error in _create_simple_video: {e}")
|
||||
return {
|
||||
"success": False,
|
||||
"error": str(e)
|
||||
}
|
||||
|
||||
async def _create_video_with_transitions(
|
||||
self,
|
||||
image_paths: List[str],
|
||||
output_path: str,
|
||||
duration_per_image: int,
|
||||
fps: int,
|
||||
transition: str
|
||||
) -> dict:
|
||||
"""
|
||||
创建带转场效果的视频
|
||||
|
||||
使用 xfade filter
|
||||
"""
|
||||
try:
|
||||
# 转场持续时间（秒）
|
||||
transition_duration = 1
|
||||
|
||||
# 构建 filter_complex
|
||||
# 每张图片显示时间 = duration_per_image
|
||||
# 转场开始时间 = duration_per_image - transition_duration
|
||||
|
||||
inputs = []
|
||||
filter_parts = []
|
||||
|
||||
# 添加所有输入
|
||||
for i, img_path in enumerate(image_paths):
|
||||
inputs.extend(['-loop', '1', '-t', str(duration_per_image), '-i', img_path])
|
||||
|
||||
# 构建 xfade filter chain
|
||||
if len(image_paths) == 2:
|
||||
# 两张图片的简单情况
|
||||
offset = duration_per_image - transition_duration
|
||||
filter_complex = f"[0][1]xfade=transition={transition}:duration={transition_duration}:offset={offset}[v]"
|
||||
else:
|
||||
# 多张图片需要链式 xfade
|
||||
current_label = "0"
|
||||
for i in range(1, len(image_paths)):
|
||||
offset = i * duration_per_image - i * transition_duration
|
||||
next_label = f"v{i}" if i < len(image_paths) - 1 else "v"
|
||||
filter_parts.append(f"[{current_label}][{i}]xfade=transition={transition}:duration={transition_duration}:offset={offset}[{next_label}]")
|
||||
current_label = next_label
|
||||
filter_complex = ";".join(filter_parts)
|
||||
|
||||
# 构建完整命令
|
||||
cmd = ['ffmpeg'] + inputs + [
|
||||
'-filter_complex', filter_complex,
|
||||
'-map', '[v]',
|
||||
'-c:v', 'libx264',
|
||||
'-pix_fmt', 'yuv420p',
|
||||
'-r', str(fps),
|
||||
'-y',
|
||||
output_path
|
||||
]
|
||||
|
||||
logger.info(f"Running FFmpeg with transitions: {' '.join(cmd[:20])}...")
|
||||
|
||||
# 执行命令
|
||||
result = subprocess.run(
|
||||
cmd,
|
||||
capture_output=True,
|
||||
text=True,
|
||||
timeout=300
|
||||
)
|
||||
|
||||
if result.returncode != 0:
|
||||
logger.error(f"FFmpeg error: {result.stderr}")
|
||||
# 如果转场失败，回退到简单模式
|
||||
logger.warning("Falling back to simple video creation")
|
||||
return await self._create_simple_video(image_paths, output_path, duration_per_image, fps)
|
||||
|
||||
# 验证输出文件
|
||||
if not os.path.exists(output_path) or os.path.getsize(output_path) == 0:
|
||||
return {
|
||||
"success": False,
|
||||
"error": "Output video file is empty or not created"
|
||||
}
|
||||
|
||||
logger.info(f"Video with transitions created successfully: {output_path}")
|
||||
return {"success": True}
|
||||
|
||||
except Exception as e:
|
||||
logger.error(f"Error in _create_video_with_transitions: {e}")
|
||||
# 回退到简单模式
|
||||
return await self._create_simple_video(image_paths, output_path, duration_per_image, fps)
|
||||
|
||||
def get_video_info(self, video_path: str) -> dict:
|
||||
"""
|
||||
获取视频信息
|
||||
|
||||
Args:
|
||||
video_path: 视频路径
|
||||
|
||||
Returns:
|
||||
dict: 视频信息
|
||||
"""
|
||||
try:
|
||||
cmd = [
|
||||
'ffprobe',
|
||||
'-v', 'quiet',
|
||||
'-print_format', 'json',
|
||||
'-show_format',
|
||||
'-show_streams',
|
||||
video_path
|
||||
]
|
||||
|
||||
result = subprocess.run(cmd, capture_output=True, text=True, timeout=10)
|
||||
|
||||
if result.returncode == 0:
|
||||
import json
|
||||
data = json.loads(result.stdout)
|
||||
|
||||
duration = float(data.get('format', {}).get('duration', 0))
|
||||
size = int(data.get('format', {}).get('size', 0))
|
||||
|
||||
return {
|
||||
"duration": duration,
|
||||
"size": size,
|
||||
"format": data.get('format', {}).get('format_name', 'unknown')
|
||||
}
|
||||
except Exception as e:
|
||||
logger.error(f"Error getting video info: {e}")
|
||||
|
||||
return {
|
||||
"duration": 0,
|
||||
"size": 0,
|
||||
"format": "unknown"
|
||||
}
|
||||
@@ -0,0 +1,134 @@
|
||||
#!/usr/bin/env python
|
||||
"""
|
||||
快速测试脚本 - 验证核心功能
|
||||
"""
|
||||
import asyncio
|
||||
import os
|
||||
import sys
|
||||
from pathlib import Path
|
||||
|
||||
# 添加路径
|
||||
sys.path.insert(0, str(Path(__file__).parent))
|
||||
|
||||
# 设置环境变量
|
||||
os.environ['OPENAI_API_KEY'] = 'sk-i9AwAgXDqqxsA9Ym4AjSPg'
|
||||
os.environ['OPENAI_BASE_URL'] = 'https://litellm.graystone-fb459c5d.southeastasia.azurecontainerapps.io/v1'
|
||||
os.environ['OUTPUT_DIR'] = './outputs'
|
||||
|
||||
from src.utils.video_processor import VideoProcessor
|
||||
from src.utils.file_manager import FileManager
|
||||
|
||||
async def test_video_processor():
|
||||
"""测试视频处理器（使用模拟图片）"""
|
||||
print("\n" + "="*60)
|
||||
print("测试: 视频处理器（FFmpeg）")
|
||||
print("="*60)
|
||||
|
||||
# 创建测试图片
|
||||
from PIL import Image
|
||||
import os
|
||||
|
||||
os.makedirs("./outputs/images", exist_ok=True)
|
||||
|
||||
image_paths = []
|
||||
colors = [(255, 0, 0), (0, 255, 0), (0, 0, 255)] # 红、绿、蓝
|
||||
|
||||
for i, color in enumerate(colors):
|
||||
img = Image.new('RGB', (1024, 1024), color)
|
||||
path = f"./outputs/images/test_image_{i}.png"
|
||||
img.save(path)
|
||||
image_paths.append(path)
|
||||
print(f"✅ 创建测试图片: {path}")
|
||||
|
||||
# 测试视频拼接
|
||||
processor = VideoProcessor()
|
||||
result = await processor.create_video_from_images(
|
||||
image_paths=image_paths,
|
||||
output_dir="./outputs/videos",
|
||||
duration_per_image=2,
|
||||
fps=30,
|
||||
transition="fade"
|
||||
)
|
||||
|
||||
if result.get("success"):
|
||||
print(f"\n✅ 视频生成成功!")
|
||||
print(f" 文件: {result['file_path']}")
|
||||
print(f" 时长: {result['duration']} 秒")
|
||||
print(f" 图片数: {result['image_count']}")
|
||||
return True
|
||||
else:
|
||||
print(f"\n❌ 视频生成失败: {result.get('error')}")
|
||||
return False
|
||||
|
||||
|
||||
async def test_file_manager():
|
||||
"""测试文件管理器"""
|
||||
print("\n" + "="*60)
|
||||
print("测试: 文件管理器")
|
||||
print("="*60)
|
||||
|
||||
manager = FileManager(base_dir="./outputs")
|
||||
|
||||
# 列出文件
|
||||
files = manager.list_files("all")
|
||||
print(f"\n📁 文件列表 (共 {len(files)} 个):")
|
||||
for file in files:
|
||||
print(f" - {file['filename']} ({file['type']}, {file['size_mb']} MB)")
|
||||
|
||||
# 存储信息
|
||||
storage = manager.get_storage_info()
|
||||
print(f"\n💾 存储信息:")
|
||||
print(f" - 总大小: {storage['total_size_mb']} MB")
|
||||
print(f" - 图片数: {storage['image_count']}")
|
||||
print(f" - 视频数: {storage['video_count']}")
|
||||
|
||||
return True
|
||||
|
||||
|
||||
async def main():
|
||||
"""运行测试"""
|
||||
print("\n" + "="*60)
|
||||
print("🧪 Video Generator Agent - 核心功能测试")
|
||||
print("="*60)
|
||||
|
||||
results = []
|
||||
|
||||
# 测试视频处理
|
||||
results.append(("视频处理器", await test_video_processor()))
|
||||
|
||||
# 测试文件管理
|
||||
results.append(("文件管理器", await test_file_manager()))
|
||||
|
||||
# 汇总
|
||||
print("\n" + "="*60)
|
||||
print("📊 测试结果")
|
||||
print("="*60)
|
||||
|
||||
passed = sum(1 for _, result in results if result)
|
||||
total = len(results)
|
||||
|
||||
for name, result in results:
|
||||
status = "✅ 通过" if result else "❌ 失败"
|
||||
print(f"{status} - {name}")
|
||||
|
||||
print(f"\n总计: {passed}/{total} 通过")
|
||||
|
||||
if passed == total:
|
||||
print("\n🎉 核心功能测试通过！")
|
||||
print("\n📝 注意: 图片生成功能需要正确的模型配置")
|
||||
print(" 当前模型 'taiji/gemini-3-pro-image-preview' 可能不支持")
|
||||
print(" 建议使用: dall-e-3, dall-e-2 或其他支持的模型")
|
||||
return 0
|
||||
else:
|
||||
return 1
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
try:
|
||||
exit_code = asyncio.run(main())
|
||||
sys.exit(exit_code)
|
||||
except Exception as e:
|
||||
print(f"\n❌ 测试失败: {e}")
|
||||
import traceback
|
||||
traceback.print_exc()
|
||||
sys.exit(1)
|
||||
@@ -0,0 +1,252 @@
|
||||
#!/usr/bin/env python
|
||||
"""
|
||||
Video Generator Agent 测试脚本
|
||||
测试图片生成和视频拼接功能
|
||||
"""
|
||||
import asyncio
|
||||
import aiohttp
|
||||
import json
|
||||
import sys
|
||||
import os
|
||||
|
||||
# 配置
|
||||
API_BASE_URL = os.getenv("API_BASE_URL", "http://localhost:8000")
|
||||
API_KEY = os.getenv("API_KEY", "sk-i9AwAgXDqqxsA9Ym4AjSPg")
|
||||
|
||||
|
||||
async def test_health_check():
|
||||
"""测试健康检查"""
|
||||
print("\n" + "="*60)
|
||||
print("测试 1: 健康检查")
|
||||
print("="*60)
|
||||
|
||||
try:
|
||||
async with aiohttp.ClientSession() as session:
|
||||
async with session.get(f"{API_BASE_URL}/health") as response:
|
||||
data = await response.json()
|
||||
print(f"✅ 状态: {response.status}")
|
||||
print(f"📊 响应: {json.dumps(data, indent=2, ensure_ascii=False)}")
|
||||
return response.status == 200
|
||||
except Exception as e:
|
||||
print(f"❌ 错误: {e}")
|
||||
return False
|
||||
|
||||
|
||||
async def test_generate_single_image():
|
||||
"""测试生成单张图片"""
|
||||
print("\n" + "="*60)
|
||||
print("测试 2: 生成单张图片")
|
||||
print("="*60)
|
||||
|
||||
try:
|
||||
payload = {
|
||||
"description": "a beautiful sunset over mountains with orange and purple sky",
|
||||
"size": "1024x1024",
|
||||
"quality": "standard"
|
||||
}
|
||||
|
||||
headers = {
|
||||
"api-key": API_KEY,
|
||||
"Content-Type": "application/json"
|
||||
}
|
||||
|
||||
print(f"📝 请求: {json.dumps(payload, indent=2, ensure_ascii=False)}")
|
||||
|
||||
async with aiohttp.ClientSession() as session:
|
||||
async with session.post(
|
||||
f"{API_BASE_URL}/api/v1/generate-image",
|
||||
json=payload,
|
||||
headers=headers,
|
||||
timeout=aiohttp.ClientTimeout(total=120)
|
||||
) as response:
|
||||
data = await response.json()
|
||||
print(f"✅ 状态: {response.status}")
|
||||
print(f"📊 响应: {json.dumps(data, indent=2, ensure_ascii=False)}")
|
||||
|
||||
if data.get("success"):
|
||||
print(f"🖼️ 图片 URL: {API_BASE_URL}{data.get('url')}")
|
||||
return True
|
||||
else:
|
||||
print(f"❌ 失败: {data.get('error')}")
|
||||
return False
|
||||
except Exception as e:
|
||||
print(f"❌ 错误: {e}")
|
||||
return False
|
||||
|
||||
|
||||
async def test_generate_video():
|
||||
"""测试生成视频"""
|
||||
print("\n" + "="*60)
|
||||
print("测试 3: 生成视频（3个场景）")
|
||||
print("="*60)
|
||||
|
||||
try:
|
||||
payload = {
|
||||
"descriptions": [
|
||||
"a peaceful morning sunrise over a calm lake with mist",
|
||||
"a busy city street at noon with people and cars",
|
||||
"a starry night sky with the milky way visible"
|
||||
],
|
||||
"duration_per_image": 3,
|
||||
"fps": 30,
|
||||
"transition": "fade"
|
||||
}
|
||||
|
||||
headers = {
|
||||
"api-key": API_KEY,
|
||||
"Content-Type": "application/json"
|
||||
}
|
||||
|
||||
print(f"📝 请求: {json.dumps(payload, indent=2, ensure_ascii=False)}")
|
||||
print("⏳ 生成视频中，这可能需要 1-3 分钟...")
|
||||
|
||||
async with aiohttp.ClientSession() as session:
|
||||
async with session.post(
|
||||
f"{API_BASE_URL}/api/v1/generate-video",
|
||||
json=payload,
|
||||
headers=headers,
|
||||
timeout=aiohttp.ClientTimeout(total=300)
|
||||
) as response:
|
||||
data = await response.json()
|
||||
print(f"✅ 状态: {response.status}")
|
||||
print(f"📊 响应: {json.dumps(data, indent=2, ensure_ascii=False)}")
|
||||
|
||||
if data.get("success"):
|
||||
video_info = data.get("video", {})
|
||||
print(f"\n🎬 视频信息:")
|
||||
print(f" - URL: {API_BASE_URL}{video_info.get('url')}")
|
||||
print(f" - 时长: {video_info.get('duration')} 秒")
|
||||
print(f" - 图片数: {video_info.get('image_count')}")
|
||||
return True
|
||||
else:
|
||||
print(f"❌ 失败: {data.get('error')}")
|
||||
return False
|
||||
except Exception as e:
|
||||
print(f"❌ 错误: {e}")
|
||||
return False
|
||||
|
||||
|
||||
async def test_list_files():
|
||||
"""测试列出文件"""
|
||||
print("\n" + "="*60)
|
||||
print("测试 4: 列出生成的文件")
|
||||
print("="*60)
|
||||
|
||||
try:
|
||||
async with aiohttp.ClientSession() as session:
|
||||
async with session.get(f"{API_BASE_URL}/api/v1/list-files?file_type=all") as response:
|
||||
data = await response.json()
|
||||
print(f"✅ 状态: {response.status}")
|
||||
|
||||
if data.get("success"):
|
||||
files = data.get("files", [])
|
||||
storage = data.get("storage", {})
|
||||
|
||||
print(f"\n📁 存储信息:")
|
||||
print(f" - 总大小: {storage.get('total_size_mb')} MB")
|
||||
print(f" - 图片数: {storage.get('image_count')}")
|
||||
print(f" - 视频数: {storage.get('video_count')}")
|
||||
|
||||
print(f"\n📄 文件列表 (共 {len(files)} 个):")
|
||||
for i, file in enumerate(files[:5], 1): # 只显示前5个
|
||||
print(f" {i}. {file['filename']} ({file['type']}, {file['size_mb']} MB)")
|
||||
|
||||
if len(files) > 5:
|
||||
print(f" ... 还有 {len(files) - 5} 个文件")
|
||||
|
||||
return True
|
||||
else:
|
||||
print(f"❌ 失败: {data.get('error')}")
|
||||
return False
|
||||
except Exception as e:
|
||||
print(f"❌ 错误: {e}")
|
||||
return False
|
||||
|
||||
|
||||
async def test_mcp_tools():
|
||||
"""测试 MCP 工具调用"""
|
||||
print("\n" + "="*60)
|
||||
print("测试 5: MCP 工具调用")
|
||||
print("="*60)
|
||||
|
||||
try:
|
||||
# 测试 tools/list
|
||||
print("\n📋 测试 tools/list:")
|
||||
payload = {
|
||||
"jsonrpc": "2.0",
|
||||
"method": "tools/list",
|
||||
"id": 1
|
||||
}
|
||||
|
||||
async with aiohttp.ClientSession() as session:
|
||||
async with session.post(
|
||||
f"{API_BASE_URL}/mcp",
|
||||
json=payload,
|
||||
timeout=aiohttp.ClientTimeout(total=30)
|
||||
) as response:
|
||||
data = await response.json()
|
||||
print(f"✅ 状态: {response.status}")
|
||||
|
||||
if "result" in data:
|
||||
tools = data["result"].get("tools", [])
|
||||
print(f"🔧 可用工具 ({len(tools)} 个):")
|
||||
for tool in tools:
|
||||
print(f" - {tool['name']}: {tool['description']}")
|
||||
return True
|
||||
else:
|
||||
print(f"❌ 失败: {data.get('error')}")
|
||||
return False
|
||||
except Exception as e:
|
||||
print(f"❌ 错误: {e}")
|
||||
return False
|
||||
|
||||
|
||||
async def main():
|
||||
"""运行所有测试"""
|
||||
print("\n" + "="*60)
|
||||
print("🧪 Video Generator Agent 测试套件")
|
||||
print("="*60)
|
||||
print(f"🌐 API URL: {API_BASE_URL}")
|
||||
print(f"🔑 API Key: {API_KEY[:20]}...")
|
||||
|
||||
results = []
|
||||
|
||||
# 运行测试
|
||||
results.append(("健康检查", await test_health_check()))
|
||||
results.append(("生成单张图片", await test_generate_single_image()))
|
||||
results.append(("生成视频", await test_generate_video()))
|
||||
results.append(("列出文件", await test_list_files()))
|
||||
results.append(("MCP 工具", await test_mcp_tools()))
|
||||
|
||||
# 汇总结果
|
||||
print("\n" + "="*60)
|
||||
print("📊 测试结果汇总")
|
||||
print("="*60)
|
||||
|
||||
passed = sum(1 for _, result in results if result)
|
||||
total = len(results)
|
||||
|
||||
for name, result in results:
|
||||
status = "✅ 通过" if result else "❌ 失败"
|
||||
print(f"{status} - {name}")
|
||||
|
||||
print(f"\n总计: {passed}/{total} 通过")
|
||||
|
||||
if passed == total:
|
||||
print("\n🎉 所有测试通过！")
|
||||
return 0
|
||||
else:
|
||||
print(f"\n⚠️ {total - passed} 个测试失败")
|
||||
return 1
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
try:
|
||||
exit_code = asyncio.run(main())
|
||||
sys.exit(exit_code)
|
||||
except KeyboardInterrupt:
|
||||
print("\n\n⚠️ 测试被中断")
|
||||
sys.exit(1)
|
||||
except Exception as e:
|
||||
print(f"\n\n❌ 测试失败: {e}")
|
||||
sys.exit(1)
|
||||
Reference in New Issue
Block a user