wtf is idea-inbox-feishu?

Idea Inbox to Feishu takes the good ideas that come up while chatting with an AI, or notes you jot down on the fly, and automatically turns them into organized content topics stored in Feishu, a Chinese team collaboration tool with spreadsheet-like multi-dimensional tables. The problem it addresses is a familiar one: you have a great idea while talking to a chatbot, then close the chat window and the idea is gone for good. This tool asks you to drop your rough notes into a folder, and lets an AI turn that mess into structured, saved topics. A key requirement is that this tool needs to be paired with an AI agent that can read files and run commands on its own, such as Claude Code, Cursor, or Codex, rather than a plain chat interface like the ChatGPT website. Most of the setup and the actual idea distillation is meant to be handled by that agent, so you typically do not need your own AI API key unless you want to turn the whole pipeline into a fully automatic scheduled script. The project sets up a Feishu base with three linked tables that form a research, content, and review workflow: a research library for judgments and supporting evidence, a content library for finished pieces like scripts and post drafts with a production status, and a review library for tracking how published content performed. These tables and their fields are all customizable through a configuration file. Day to day use involves saving notes as plain text or markdown files into an inbox folder, running a scan command to list what needs processing, having an AI read that scan output and turn it into structured records using a provided prompt template, then pushing those records into the correct Feishu table and archiving the original files. The tool is built with Python's standard library only, so no extra packages need to be installed, though a separate command line tool for Feishu is required to actually write into the tables. It is aimed at content creators, researchers, and anyone who regularly has useful ideas surface during AI conversations. The project is released under the MIT license.