为啥网页只是readme上的东西,还有这个readme不要出现这个内容wiki/graph-data.json and wiki/knowledge-gr…

Metadata

  • Stable ID: codex-user-prompt:9fed3741f5dc3052
  • Source kind: codex-session-user
  • Category: wiki-ingest
  • Timestamp: 2026-05-10T21:04:57.310Z
  • Semantic hash: 9fed3741f5dc3052c8c625d4edee48d739eaf759b19bf72b2aa27ccd0acc5d32
  • Public handling: selected full-text prompt with secret filtering.

Prompt Text

为啥网页只是readme上的东西,还有这个readme不要出现这个内容wiki/graph-data.json and wiki/knowledge-graph.html can still be generated, but graph output is secondary to ingest quality, search quality, and durable retrieval.
 
现在网页做的还是太toy了,你可以借鉴主流的这方面的思想,比如多个开源项目,看看他们的网页制作,找找灵感,多agent协作实现一个,互动性强,关联性强,实用性强的网页
 
同时关于做项目和科研,我想加一条说法,让agent记住,就是每当我说比较抽象比较general的内容和词语的时候,你就要广泛深入的去多个相关主流github项目,顶会论文项目那边找灵感,汲取经验,我们不抄袭,但是我们需要知识内容

Reuse Notes

  • EXTRACTED: This is a selected Codex prompt or automation prompt from the local Codex corpus.
  • INFERRED: Future agents can reuse its structure, constraints, and acceptance criteria when creating similar Codex workflows.