费曼工作流 是 AI Skill Hub 本期精选Agent工作流之一。综合评分 7.5 分,整体质量较高。我们推荐使用将其纳入你的 AI 工具库,帮助提升工作效率。
费曼工作流 是一套完整的 AI Agent 自动化工作流方案。通过可视化的节点编排,将复杂的多步骤任务拆解为清晰的自动化流程,实现全程无人值守的智能处理。支持与数百种外部服务和 API 无缝集成,适合构建数据处理管线、业务自动化和 AI 辅助决策系统。
费曼工作流 是一套完整的 AI Agent 自动化工作流方案。通过可视化的节点编排,将复杂的多步骤任务拆解为清晰的自动化流程,实现全程无人值守的智能处理。支持与数百种外部服务和 API 无缝集成,适合构建数据处理管线、业务自动化和 AI 辅助决策系统。
# 方式一:pip 安装(推荐)
pip install feynman
# 方式二:虚拟环境安装(推荐生产环境)
python -m venv .venv
source .venv/bin/activate # Windows: .venv\Scripts\activate
pip install feynman
# 方式三:从源码安装(获取最新功能)
git clone https://github.com/steveyeow/Feynman
cd Feynman
pip install -e .
# 验证安装
python -c "import feynman; print('安装成功')"
# 命令行使用
feynman --help
# 基本用法
feynman input_file -o output_file
# Python 代码中调用
import feynman
# 示例
result = feynman.process("input")
print(result)
# feynman 配置文件示例(config.yml) app: name: "feynman" debug: false log_level: "INFO" # 运行时指定配置文件 feynman --config config.yml # 或通过环境变量配置 export FEYNMAN_API_KEY="your-key" export FEYNMAN_OUTPUT_DIR="./output"
"You learn by asking questions, by thinking, and by experimenting." — Richard Feynman
Chat with books. Great minds join in.
An interactive knowledge network built on the world's most important books and great minds. With Feynman, you can chat with the books you want to read to quickly understand them and explore the broader context around them. You can also start from a topic, and Feynman will surface the most relevant books to help you build a knowledge system grounded in them. As you chat, a continuously evolving network of agent-simulated great minds — scholars, scientists, practitioners — automatically join the conversation, so you read, learn, and discuss ideas together with the most relevant thinkers. Books, minds, and ideas are all interconnected nodes in this network — not isolated features, but parts of a navigable map of human thought.
What makes this different:
This AI-generated overview summarizes the cognitive science behind the method (Russian Activity Theory and Chinese metacognitive scaffolding) and explains the Pre-Thinking Stage that frames the learning flow. Watch the video here: Feynman Reading Method Overview.
Every LLM call shows its token consumption in real time — chat, discovery, and search. No hidden costs.
git clone https://github.com/steveyeow/feynman.git
cd feynman
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
cp .env.example .env # add at least one API key
uvicorn app.main:app --reload
Open http://localhost:8000
Edit .env to set your LLM provider keys. At least one is required:
| Variable | Provider | Notes |
|---|---|---|
GEMINI_API_KEY | Google Gemini | Recommended — supports embeddings + web search grounding |
DEEPSEEK_API_KEY | DeepSeek | Cost-effective chat, no embeddings |
OPENAI_API_KEY | OpenAI | GPT-4o-mini for chat, text-embedding-3-small for embeddings |
KIMI_API_KEY | Moonshot Kimi | Chat only, no embeddings |
ANTHROPIC_API_KEY | Anthropic Claude | Chat only, no embeddings |
The system auto-selects the best available provider and falls back through the chain: Gemini → DeepSeek → OpenAI → Kimi → Anthropic.
| Variable | Default | Description |
|---|---|---|
VOTE_THRESHOLD | 3 | Upvotes needed to auto-index a book |
DISCOVERY_INTERVAL | 21600 | Seconds between scheduled discovery runs (0 to disable) |
DISCOVERY_BATCH_SIZE | 5 | Max books discovered per scheduled run |
TOPIC_DISCOVER_COUNT | 5 | Books discovered per topic click |
创新性的AI工作流项目
该工具使用 NOASSERTION 协议,商用场景请仔细阅读协议条款,必要时咨询法律意见。
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
📄 NOASSERTION — 请查阅原始协议条款了解具体使用限制。
经综合评估,费曼工作流 在Agent工作流赛道中表现稳健,质量良好。如果你已有明确的使用需求,可以直接上手体验;如果还在评估阶段,建议对比同类工具后再做决策。
| 原始名称 | Feynman |
| Topics | aieducationfastapipython |
| GitHub | https://github.com/steveyeow/Feynman |
| License | NOASSERTION |
| 语言 | Python |
收录时间:2026-05-27 · 更新时间:2026-05-27 · License:NOASSERTION · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端