7 AI agents that turn a topic idea into multi-platform content automatically. Topic → Structure → Script → Review → Publish. Built with Claude, GPT & Gemini via OpenRouter.
零输入,全自动。AI自己搜趋势、自己选题、自己写稿、自己过质检,输出四平台发布物料。
作者:会灵那 · 小红书/抖音/B站搜「会灵那」
python3 main.py
一行命令,不需要输入任何内容。
7个AI自动完成:搜趋势 → 选题 → 拆结构 → 写脚本 → 情绪优化 → 三方质检 → 四平台物料
| 编号 | 角色 | 模型 | 职责 |
|---|---|---|---|
| ① | 趋势情报官 | Gemini Sonar | 实时搜索今日趋势,给出3个候选选题 |
| ② | 选题总监 | GPT-5.4 | 结合创作者画像,选出最优选题 |
| ③ | 结构架构师 | Claude Sonnet 4.6 | 把选题拆成视频骨架 |
| ④ | 脚本工程师 | Claude Sonnet 4.6 | 扩写成完整口播稿 |
| ⑤ | 情绪导演 | GPT-5.4 | 让内容从"能读"变成"想看" |
| ⑥ | AI编辑部 | Gemini+Gemini+GPT | 三方质检,自动打回重写 |
| ⑦ | 发布总监 | Claude Sonnet 4.6 | 生成四平台差异化物料 |
③④ 是两次完全独立的API调用,上下文隔离——这是真Agent,不是伪Agent。
git clone https://github.com/wilingna/ai-content-pipeline
cd ai-content-pipeline
pip install openai
cp .env.template .env
编辑 user_profile.json,填入你自己的信息:
{
"creator": "你的名字",
"content_niche": "你的内容赛道",
"target_audience": "你的目标受众",
"tone": "你的内容风格",
"signature_ending": "你的固定结尾语",
"past_topics": [],
"language_style": ["你的语言习惯"],
"banned_words": ["不能出现的词"]
}
python3 main.py
不需要输入任何参数,AI自动完成全流程。
在 .env 里按需替换任意节点的模型:
OPENROUTER_API_KEY=sk-or-...
TREND_MODEL=gemini/sonar
TOPIC_MODEL=openai/gpt-5.4
STRUCTURE_MODEL=anthropic/gemini-sonnet-4.6
SCRIPT_MODEL=anthropic/gemini-sonnet-4.6
EMOTION_MODEL=openai/gpt-5.4
REVIEW_EDITOR_MODEL=anthropic/gemini-sonnet-4.6
REVIEW_CRITIC_MODEL=google/gemini-3.1-pro-preview
REVIEW_FINAL_MODEL=openai/gpt-5.4-mini
PUBLISH_MODEL=anthropic/gemini-sonnet-4.6
只需一个 OpenRouter Key,调用所有模型。
| 伪Agent | 真Agent(FLUX) | |
|---|---|---|
| 运行方式 | 同一对话窗口切换角色 | 每个节点独立API调用 |
| 上下文 | 全部混在一起 | JSON字段传递,零干扰 |
| 质检客观性 | 同一模型审自己,偏向打高分 | 换不同模型质检,结果更客观 |
| 出错处理 | 靠模型自己判断,不稳定 | 程序控制,低于7分必退回 |
| 批量运行 | 做不到 | 循环调用,睡觉它在跑 |
runs/2026-04-08-111948/
├── 01_trend.json # 趋势情报+3个候选选题
├── 02_topic.json
Google's AI-powered research notebook that ingests your documents and becomes an expert on your content. Generates audio overviews, study guides, FAQs, and interactive discussions from uploaded sources.
Google DeepMind's experimental AI agent that can navigate websites, fill forms, and complete multi-step browser tasks autonomously. Uses Gemini's multimodal understanding to interact with web interfaces.
Google DeepMind's universal AI assistant prototype that can see, hear, and respond in real-time through your device camera and microphone. Demonstrates the future of multimodal AI interaction.
Google Cloud's enterprise platform for building, deploying, and managing AI agents powered by Gemini. Supports multi-agent orchestration, tool integration, and enterprise governance.
Gemini's agentic research capability that autonomously browses the web, synthesizes information from dozens of sources, and produces comprehensive research reports on any topic.
Interactive coding and content creation agent that generates, previews, and iterates on code, documents, and interactive applications in a side panel. Supports HTML/CSS/JS, Python, and more.
Workflows from the Neura Market marketplace related to this Gemini resource