LLM Wiki
Brain app
First committed by jackwener
Published in jackwener/llm-wiki. Publisher page
107 GitHub stars
What it is
Agent-native LLM Wiki — AI-maintained knowledge base with Obsidian compatibility
AutoGineer overview, from the brain’s own files
LLM Wiki is a CLI-driven framework designed to transform raw documents into a structured, interconnected knowledge base. Unlike traditional RAG systems that re-process raw data, this tool uses AI agents to compile information into persistent wiki pages. It provides a standardized vault structure, including bootstrap files for AI agents and operation-specific skills for ingesting, querying, linting, and researching. The system is designed to be human-readable via Obsidian, while remaining fully operational for AI agents through slash commands.
How it works
Memory model: LLM wiki
AutoGineer overview, from the brain’s own files
The system initializes a local vault containing a wiki directory for structured content and a sources directory for raw documents. It uses agent-specific bootstrap files (CLAUDE.md and AGENTS.md) to guide AI agents in managing the vault. Four core skills (ingest, query, lint, research) are installed into the agent's workspace, allowing the agent to read sources, update wiki pages, perform health checks, and conduct external research. A CLI tool manages the vault, provides graph analysis, and supports optional vector search via DB9 integration.
| Folder | Role | Files |
|---|---|---|
| skills | agent-instructions | 4 |
| src | source-code | 14 |
| test | tests | 7 |
Requirements
Runtimes: Node.js >=20
Databases: PostgreSQL
Works with: Claude Code, Claude apps, OpenAI Codex, GitHub Copilot, Cursor, Gemini CLI, Google Antigravity, Kiro, OpenCode, Amp, Roo Code, Cline, JetBrains Junie, Devin (ex-Windsurf)
Set it up
- Not scanned by AutoGineerAutoGineer could not scan everything it runs. https://www.npmjs.com/package/@jackwener/llm-wiki
First committed by jackwener. Published in jackwener/llm-wiki. Licence: Apache License 2.0.
Usage
AutoGineer overview, from the brain’s own files
- Initialize a new wiki vault in a local directory.
- Use AI agents to ingest raw documents into structured wiki pages.
- Query the wiki to synthesize answers and compound knowledge.
- Perform automated health checks on the wiki graph.
- Conduct external research and integrate findings into the wiki.
- Analyze the wikilink graph to identify clusters, hubs, and orphans.
- “/ingest sources/project-notes.md”
- The agent reads the source file, extracts key entities, and updates or creates corresponding wiki pages with wikilinks.
- “/query "What is the current status of the API migration?"”
- The agent searches the wiki, synthesizes an answer based on existing pages, and writes back new insights to the wiki.
- “/lint”
- The agent performs a health check to identify broken links, orphans, and contradictions, and auto-fixes safe issues.
- “/research "latest trends in edge computing"”
- The agent searches the web, saves relevant sources, ingests them into the wiki, and synthesizes a report.
Documentation by jackwener
LLM Wiki
Agent-native persistent knowledge management — compile knowledge once, query forever.
Based on Andrej Karpathy's LLM Wiki pattern.
What is this?
LLM Wiki is a CLI tool + AI Agent skill system that maintains an evolving, interconnected Markdown knowledge base. Instead of traditional RAG (re-deriving answers from raw documents each time), LLM Wiki compiles knowledge into structured wiki pages that AI agents maintain and grow over time.
Key principle: The tool itself doesn't call LLMs. It provides skill files that let any AI agent (Claude Code, Codex, etc.) operate the wiki. Obsidian is the human interface — no self-built GUI.
Quick Start
# Install globally
npm install -g @jackwener/llm-wiki
# Initialize a new wiki vault
mkdir my-wiki && cd my-wiki
llm-wiki init
# Now use your AI agent:
# /ingest sources/some-article.md
# /query "What do we know about X?"
# /lint
# /research "deep dive on Y"llm-wiki init is the only setup command — it creates the vault files, the
agent bootstrap files (CLAUDE.md, AGENTS.md), and installs the four bundled
operation skills into .claude/skills/ and .agents/skills/ in one step.
After upgrading the package, refresh the installed skill files with:
llm-wiki skill installVault Structure
my-wiki/
├── CLAUDE.md # Agent bootstrap for Claude Code (auto-loaded)
├── AGENTS.md # Agent bootstrap for Codex (auto-loaded)
├── wiki-purpose.md # Wiki scope and audience
├── wiki-schema.md # Page types, naming conventions, frontmatter rules
├── wiki-log.md # Append-only operation log
├── wiki/ # AI-maintained wiki pages (Obsidian-compatible)
├── sources/ # Raw, immutable source documents
│ └── YYYY-MM-DD/ # Date-based storage
├── .claude/
│ └── skills/
│ ├── ingest/SKILL.md # Source → wiki pages
│ ├── query/SKILL.md # Evidence-grounded wiki answers
│ ├── lint/SKILL.md # Wiki health checks
│ └── research/SKILL.md # External research → wiki
├── .agents/
│ └── skills/
│ ├── ingest/SKILL.md
│ ├── query/SKILL.md
│ ├── lint/SKILL.md
│ └── research/SKILL.md
└── .llm-wiki/
├── config.toml # Vault configuration
└── sync-state.json # Incremental sync trackingllm-wiki init generates every file above in one step.
Agent Bootstrap
LLM Wiki uses a two-file pattern so any AI agent can operate the vault out of
the box, with no manual setup beyond llm-wiki init:
1. Entry files — CLAUDE.md and AGENTS.md (vault root)
Short bootstrap documents that are auto-loaded on every session start —
Claude Code reads CLAUDE.md, Codex reads AGENTS.md. They tell the agent:
this workspace is an LLM Wiki vault
where to find
wiki-purpose.mdandwiki-schema.mdwhich
/ingest,/query,/lint,/researchcommands are availablewhich of the four operation skills to load (
ingest,query,lint, orresearch)a short CLI cheat-sheet and the core operating rules
Because they are auto-loaded, they are intentionally small — a few dozen lines — to keep session-start context cheap.
2. Operation skills — .claude/skills/{ingest,query,lint,research}/SKILL.md
and .agents/skills/{ingest,query,lint,research}/SKILL.md
Each operation has a focused playbook, loaded on demand when the matching
command is invoked. The skills contain the relevant procedure, page schemas,
frontmatter rules, and invariants (for example, source immutability and the
wiki-log.md + sync tail rule). Every skill follows the
Agent Skills specification
(skills/<name>/SKILL.md with YAML frontmatter), so spec-compliant agents
(Claude Code, Codex, Amp, and others) discover them independently. All four
are installed into both platform directories for a shared vault setup.
Upgrading. llm-wiki init is the only setup command — it writes both
entry files and installs the four skills. After upgrading the npm package, run
llm-wiki skill install to refresh the skill files. Your edits to
CLAUDE.md / AGENTS.md are preserved across reinstalls.
Upgrading to the four-skill layout. Run llm-wiki skill install to add
the operation skills. If you previously installed the legacy monolithic skill,
remove it afterwards so its broad trigger does not overlap the four focused
skills:
rm -rf .claude/skills/llm-wiki .agents/skills/llm-wikiOperations
The skill exposes four operations, each invoked as a slash command:
| Operation | Usage | What it does |
|---|---|---|
| ingest | /ingest <path> | Read source → extract entities → create/update wiki pages with [[wikilinks]] |
| query | /query <question> | Search wiki → synthesize answer → write back valuable insights (knowledge compounding) |
| lint | /lint | Health check: broken links, orphans, contradictions, stale content → auto-fix safe issues |
| research | /research <topic> | Go beyond wiki: search web → save sources → ingest → synthesize report |
CLI Commands
| Command | Description |
|---|---|
llm-wiki init [dir] | Initialize a new wiki vault |
llm-wiki search <query> | BM25 keyword search (+ DB9 vector search if configured) |
llm-wiki graph [--json] | Analyze wikilink graph: communities, hubs, orphans, wanted pages |
llm-wiki status | Wiki statistics and health summary |
llm-wiki sync [--dry-run] | Track changes (mtime + SHA256), sync embeddings to DB9 |
llm-wiki skill install | Install all skills to your AI agent workspace |
llm-wiki skill list | List available skills |
llm-wiki skill show <name> | Print skill content to stdout |
Search
BM25 keyword search with CJK bigram tokenization (Chinese/Japanese/Korean support).
When DB9 is configured, search becomes hybrid: BM25 + vector similarity, merged via Reciprocal Rank Fusion (RRF, K=60).
llm-wiki search "distributed consensus"
llm-wiki search "分布式共识" -n 5
llm-wiki search "raft algorithm" --bm25-onlyGraph Analysis
Analyzes the [[wikilink]] graph to find structure in your knowledge:
Communities — Topic clusters detected via label propagation
Hub pages — Most connected pages (high incoming + outgoing links)
Orphan pages — Pages with no incoming links
Wanted pages — Pages linked but not yet created
llm-wiki graph # Human-readable output
llm-wiki graph --json # Machine-readable for programmatic useDB9 Integration (Optional)
DB9 adds vector search and cloud sync:
Server-side embeddings via
embedding(text)::vector(1024)— no local model neededHNSW vector index for semantic similarity search
Reverse source lookup: "which wiki pages reference this source?"
Enable by adding to .llm-wiki/config.toml:
[db9]
url = "your-db9-connection-string"Then run llm-wiki sync to upload embeddings.
Obsidian Compatibility
The wiki/ directory is a standard Obsidian vault:
YAML frontmatter
[[wikilink]]cross-referencesOpen directly in Obsidian for browsing, graph view, and editing
Configuration
.llm-wiki/config.toml:
[vault]
name = "My Wiki"
language = "en"
# Optional: DB9 for vector search + cloud sync
# [db9]
# url = "your-db9-connection-string"Tech Stack
TypeScript (ESM, Node 20+)
Commander.js — CLI framework
gray-matter — Frontmatter parsing
pg — PostgreSQL client (for DB9)
tsup — Build
Vitest — Testing
License
Apache-2.0
LICENSE
Shown as plain text.
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work.
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to the Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by the Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding any notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
Copyright 2024 LLM Wiki Contributors
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
34 files · 31 not served here
- .gitignore (not served here)
- ATTRIBUTION.md
- LICENSE
- README.md
- package-lock.json (not served here)
- package.json (not served here)
- skills/ingest/SKILL.md (not served here)
- skills/lint/SKILL.md (not served here)
- skills/query/SKILL.md (not served here)
- skills/research/SKILL.md (not served here)
- src/cli.ts (not served here)
- src/commands/graph.ts (not served here)
- src/commands/init.ts (not served here)
- src/commands/search.ts (not served here)
- src/commands/skill.ts (not served here)
- src/commands/status.ts (not served here)
- src/commands/sync.ts (not served here)
- src/lib/config.ts (not served here)
- src/lib/db9.ts (not served here)
- src/lib/graph.ts (not served here)
- src/lib/search.ts (not served here)
- src/lib/skills.ts (not served here)
- src/lib/sync.ts (not served here)
- src/lib/wiki.ts (not served here)
- test/db9.test.ts (not served here)
- test/graph.test.ts (not served here)
- test/init.test.ts (not served here)
- test/rrf.test.ts (not served here)
- test/search.test.ts (not served here)
- test/sync.test.ts (not served here)
- test/wiki.test.ts (not served here)
- tsconfig.json (not served here)
- tsup.config.ts (not served here)
- vitest.config.ts (not served here)
Advantages
- Compiles knowledge into structured pages rather than re-deriving answers from raw documents, leading to knowledge compounding. evidence: README.md
- Provides native Obsidian compatibility for human browsing, graph visualization, and manual editing. evidence: README.md
Limitations
- Requires manual removal of legacy skill files if upgrading from a monolithic skill layout. evidence: README.md
- Vector search and cloud sync capabilities require an external DB9 configuration. evidence: README.md
Security
Risk level: No risk found
This brain runs
Agent instructions that run or fetch something (5)
skills/ingest/SKILL.md:Run `llm-wiki sync` after the changes.skills/lint/SKILL.md:Run `llm-wiki sync` when changes were made.skills/query/SKILL.md:Link it to the contributing pages, append a `query` entry to `wiki-log.md`, and run `llm-wiki sync`.skills/research/SKILL.md:Run the `query` workflow first to determine what is known and define a precise research question.skills/research/SKILL.md:Then append a `research` entry to `wiki-log.md` and run `llm-wiki sync`.
Attribution and licence
- First committed by
jackwener
- Publisher
- jakevin (Community)
- Source
- jackwener/llm-wiki · commit cd54666
- Licence
- Apache License 2.0 (files served here)
- ATTRIBUTION.md (written by AutoGineer)
- Synced
- Contributors
- jackwener
AutoGineer is not affiliated with or endorsed by jackwener.
Install from source
31 of its files are not served here, so there is no ZIP. jackwener/llm-wiki at cd54666