菜单
File new articles, transcripts, and notes into knowledge/ in the right place with source metadata and an evidence tag.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
Evaluates agent output adversarially before committing — prevents self-assessment bias
Writes a structured session handoff so a fresh agent can continue without losing thread
Turn a plain-English vision or task into a verified, build-ready spec before any code. Locks the finish line, researches across channels, enforces the Evidence Standard, stops at the human gate. Adopted from CleanExpo/Fabel-Prompt-Engineer for the Authority-Site / Pi-CEO / Unite-Group build process.
Loop through advisor profiles in knowledge/board/ and return a combined multi-persona critique of a spec, draft, or decision.
Turn a plain-English vision into a verified, build-ready spec. Locks the finish line, researches across channels, enforces the Evidence Standard, and stops at the human approval gate.
Capture the user's feedback about an output and update the relevant skill or knowledge files so future outputs get sharper.
| name | ingest |
| description | File new articles, transcripts, and notes into knowledge/ in the right place with source metadata and an evidence tag. |
Input: a new resource — article, transcript, note, Obsidian export, or a URL.
Classify the resource:
knowledge/board/<advisor>/ (create the
advisor folder + profile.md stub if new).knowledge/notes/.knowledge/frameworks/.File it as Markdown with a metadata header:
---
source: <URL, "user", or original filename>
ingested: <YYYY-MM-DD>
evidence: <verified | unconfirmed> # verified only if the source is checkable
topic: <few words>
---
Don't silently rewrite. Preserve the original content; add structure (headings, speaker labels for transcripts) only where it aids retrieval.
Report back: where it was filed, under what evidence tag, and which
skills will now pick it up (e.g. a new advisor becomes visible to
ask-the-board).
When given a zipped Obsidian vault or a batch of transcripts: propose a filing plan first (counts per destination folder), get approval, then ingest. Flag anything that looks like a secret or personal credential instead of filing it.