Skip to main content

内容分发工作流 v0.1 · 抓取英文营销内容 → 改写成公众号/小红书中文草稿

Project description

aether

内容分发工作流 v0.1 · MVP 验证期 · LangGraph day-1 编排

一句话

把英文优质内容自动抓回来,按微信公众号 + 小红书双平台风格改写成中文草稿,进入人工审核 / 发布。 第一阶段:单人自用,验证「双平台适配 + 降 AI 味」的核心假设。真实平台自动发布留到 v0.2+

一键安装(macOS 用户)

brew install postgresql@16 && brew services start postgresql@16 && createdb aether
pip install aether-mkt
aether init      # 自动建表 + 自动下 Playwright chromium(~95MB 一次性)+ 引导填 LLM_API_KEY
aether serve     # launchd 后台跑 web + 自动重启
# 浏览器 → http://localhost:8000/articles

aether init 幂等——重跑不重复建表/下浏览器。被 Cloudflare / JS 渲染挡的源需要 Playwright 的 rendered fetcher,浏览器二进制走 init 自动装;离线/防火墙装不下来 init 也不会中断,其它源照样能跑。

CLI 命令:aether {init,serve,stop,update,status,logs,uninstall} 完整说明:docs/install-mac.md

业务领域

  • 内容来源:英文营销生态(社媒营销、内容营销、SEO、Google ads、综合营销)——见 blocks/sources/basic/sources.yaml
  • 目标平台:微信公众号(深度长文) + 小红书(图文卡片)
  • 核心差异:降 AI 味 + 按平台适配的语言风格改写,不是单纯翻译

路线

按块验证,从最小可验证单元开始。LangGraph 骨架提前铺,每个 block = 1 个 subgraph,独立测过再接主图。

详见 docs/roadmap.md

当前活跃块:Block 0 · 数据源(最小)

技术栈

  • Python 3.11+
  • LangGraph(主图编排,Day 1 起铺)
  • Anthropic SDK(Block 3+ 的 LLM 调用)
  • PostgreSQL 16(运行时状态库,详见 ADR-0004)
  • pyyaml(sources 配置)+ selectolax / httpx(Block 0 解析)
  • pytest(测试)

详见 pyproject.tomldocs/tech-stack.md

文档

我想知道 去看
整个架构设计 docs/design.md
架构图(手稿的数字版) docs/architecture.html
现在做到哪、接下来做什么 docs/roadmap.md
已经拍了哪些决策 docs/decisions/
每一块在验证什么 blocks/
每次验证的判断记录 logs/
数据源白名单 blocks/sources/basic/sources.yaml

协作方式

你是「开发者 + 决策者 + 第一个用户」。AI agent 是协作搭子,不替你拍板

风格 prompt 不能漂 —— 都进 git,可回滚可对比。

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

aether_mkt-0.3.2.tar.gz (465.4 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

aether_mkt-0.3.2-py3-none-any.whl (266.2 kB view details)

Uploaded Python 3

File details

Details for the file aether_mkt-0.3.2.tar.gz.

File metadata

  • Download URL: aether_mkt-0.3.2.tar.gz
  • Upload date:
  • Size: 465.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.11

File hashes

Hashes for aether_mkt-0.3.2.tar.gz
Algorithm Hash digest
SHA256 31b3b47cfd745dc14896f41ecb2f90a0bb4910a69c4969cc851323b41a6d8595
MD5 bc1dfec37bb9c3826fa696445dbae2d0
BLAKE2b-256 eaee79c857f16de574a81c915d580864d9cd9f7c59965ae63a444c69dae3b699

See more details on using hashes here.

File details

Details for the file aether_mkt-0.3.2-py3-none-any.whl.

File metadata

  • Download URL: aether_mkt-0.3.2-py3-none-any.whl
  • Upload date:
  • Size: 266.2 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.11

File hashes

Hashes for aether_mkt-0.3.2-py3-none-any.whl
Algorithm Hash digest
SHA256 05c521c6d961afe362eeea9aad8d806bed748f515d9e50a940a970b04201ae13
MD5 db9232ba30de4076b0b2ea780b1bdf1b
BLAKE2b-256 0ab124e05f7547ac43b275fbd3089b93a3efe747afff1ff771e463f1c705395e

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page