A knowledge-base compiler that turns source notes and created work into summaries, concepts, and indexes stored in a wiki folder. It distinguishes outside material from the user’s own work.

Use it to process articles, books, podcasts, papers, notes, projects, or artifacts into structured wiki entries. It creates dated summaries and updates concept pages with sources and relationships.

It removes the manual work of organizing new material and connecting it with an existing knowledge base. It can also identify files that have not yet been compiled.

How gatelynch/llm-knowledge-base configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to gatelynch/llm-knowledge-base. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/gatelynch/llm-knowledge-base/main/.claude/commands/compile.md

Clone the repo

git clone --depth 1 https://github.com/gatelynch/llm-knowledge-base

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/commands/gatelynch/llm-knowledge-base/compile/github.svg)](https://agentmods.dev/commands/gatelynch/llm-knowledge-base/compile)

Your own site

<a href="https://agentmods.dev/commands/gatelynch/llm-knowledge-base/compile"><img src="https://agentmods.dev/badge/commands/gatelynch/llm-knowledge-base/compile/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/commands/gatelynch/llm-knowledge-base/compile"><img src="https://agentmods.dev/badge/commands/gatelynch/llm-knowledge-base/compile.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 23 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 1,372 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00023	$0.01372
Opus 5.5	$0.00009	$0.00549
Sonnet 5.5	$0.00005	$0.00274
Haiku 4.5	$0.00002	$0.00137

Measured 2mo ago against content hash b905c6a7f2f0, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

compile scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/commands/compile.md · 161 lines

How it starts

The opening of the file, as written. The whole thing — 161 lines — stays where its author put it; the contents beside it link to each section on GitHub.

知識庫編譯

把 raw/（素材）與 artifacts/（你的作品）編譯成結構化知識，存入 wiki/。

流程

1. 確認編譯範圍

如果使用者指定了路徑或數量，就處理那個範圍
如果沒有指定，讀取 wiki/indexes/All-Sources.md，與 raw/ 和 artifacts/ 比對，找出尚未編譯的新檔案
如果新檔案超過 15 個，告知使用者，建議分批處理（每批 10-15 個）

2. 判斷來源類型

根據檔案的子資料夾路徑自動判斷 origin：

origin: external（外部來源——他人撰寫）

raw/articles/
raw/books/
raw/podcasts/
raw/papers/

origin: self（自己的——使用者撰寫）

raw/notes/
raw/projects/
artifacts/*

3. 讀取每個來源檔案

讀取每個檔案並理解其內容。依日期由新到舊處理。

4. 生成摘要（wiki/summaries/）

每個來源產出一份摘要。檔名格式：YYYYMMDD 簡短標題.md

外部來源（origin: external）

---
origin: external
source: "[[原始檔案名稱]]"
compiled: YYYY-MM-DD
tags: [tag1, tag2]
---

# 標題

## 核心結論
（1-3 句話——最重要的收穫）

## 關鍵證據
（支持結論的具體事實、數據、引述）

## 疑點
（不確定、有爭議或尚未驗證的主張）

## 術語
（這個來源引入的重要術語，附簡短定義）

自己的作品（origin: self）

---
origin: self
source: "[[你的作品檔名]]"
compiled: YYYY-MM-DD
tags: [tag1, tag2]
---

# 標題

## 我的主張
（這篇作品的核心論點或目標）

## 實踐經驗
（做了什麼、成效如何、實際發生了什麼）

## 未解問題
（寫完或做完後仍懸著的問題）

## 與研究的對照
（你的經驗如何與已知研究吻合或矛盾）

5. 提取或更新概念（wiki/concepts/）

從摘要中找出概念。檔名格式：概念名稱.md

新概念：建立條目

---
concept: 概念名稱
related: [相關概念1, 相關概念2]
updated: YYYY-MM-DD
sources:
  - "[[來源1]]"
---

# 概念名稱

## 定義

## 我的實踐
（從 origin: self 來源整理——如何使用這個概念、發生了什麼、學到什麼）

## 外部觀點
（從 origin: external 來源整理——研究怎麼說、他人如何定義）

## 張力與缺口
（你的經驗與外部研究之間的矛盾，或尚未驗證的地方）

## 例子

## 來源
### 我的
### 外部

現有概念的更新規則：

在 sources 清單中附加新來源
依 origin 把新內容路由到對應段落：origin: self → 「我的實踐」，origin: external → 「外部觀點」
如果概念同時有 self 和 external 來源，檢查「張力與缺口」是否需要更新
在「例子」中新增新例子
更新 updated 日期
在來源清單中將來源歸類至「我的」或「外部」

6. 更新索引（wiki/indexes/）

All-Sources.md：為每個新編譯的來源新增一行（來源、標籤、核心收穫、狀態）
All-Concepts.md：如果有新建概念條目，新增一行（概念、條目連結、定義、相關概念）

7. 回報結果

告訴使用者：

共編譯了多少個檔案（分 external 與 self）
建立了哪些摘要
建立或更新了哪些概念
索引更新狀態

編譯原則

不修改 raw/ 或 artifacts/：素材是唯讀的
source 是必填欄位：每份摘要都必須有 wikilink 指回原始檔案
origin 是必填欄位：external 或 self，從來源路徑自動判斷
疑點不能留空（external）：如果沒有發現問題，寫「這個來源的論點有充分支撐；沒有發現重大疑點」
概念需要交叉引用：只有在 2 個以上摘要中出現的概念才建立獨立條目。單一提及放在摘要的「術語」段落即可
檔名不加前綴：摘要與概念的檔名不加 S-、C- 等前綴
概念條目以使用者視角為主：「我的實踐」在「外部觀點」之前

Read the full file on GitHub · 161 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

2mo ago First seen · 161 lines · 23 tokens per session scan A b905c6a7f2f0

Subscribe to this mod's changes

compile is a command published in the GitHub repository gatelynch/llm-knowledge-base (343 stars, last pushed 7mo ago), licensed MIT. It adds 23 tokens to every session and 1,372 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Documents Writing and docs

compile next to 6 commands from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
compilegatelynch/llm-knowledge-base	Command	343	7mo ago	MIT	A	23	Claude Code
brand-generateferdinandobons/brand-docs	Command	291	3mo ago	MIT	A	0	Claude Code
cti-report7onez/cti-expert	Command	622	15d ago	none	A	45	Claude Code
render-figuresbraininahat/brains-in-a-hat	Command	4	18d ago	MIT	A	17	Claude Code
ppt-image2-editable-rebuildwwe-dog/ppt-image2-editable-rebuild	Command	204	5mo ago	Unlicense	A	15	Claude Code
publish-api-docsnavraj007in/architecture-cowork-plugin	Command	3	4mo ago	Apache-2.0	A	17	Claude Code
write-patentjiutuhky/my-super-capsule	Command	11	8mo ago	MIT	A	48	Claude Code

llm-knowledge-base: Command for Claude Code