- Renamed sections in the wiki index from Entities/Sources/Concepts to a more structured format. - Removed the log.md file and its associated tests to streamline the logging process. - Updated the ai-agent-patterns.md to include JSON write examples and demo conventions. - Modified commands tests to handle JSON input for writing and reading pages. - Implemented delete functionality for pages and ensured proper index updates. - Enhanced index management to support upserting entries and handling duplicates. - Removed deprecated profile tests and log manager tests to clean up the codebase. - Adjusted storage tests to reflect changes in file writing locations.
7.8 KiB
llmwiki-cli
A CLI tool for LLM agents to build and maintain personal knowledge bases.
Inspired by Andrej Karpathy's LLM Wiki.
Live Demo — interactive d3-force graph built from the example wiki in test-wiki-page/.
Overview
The CLI is the hands -- it reads, writes, searches, and manages wiki files. The LLM is the brain -- it decides what to create, update, and connect.
LLM Agent (Claude Code / Codex)
|
| shells out to:
| $ wiki init my-wiki --name "Notes" --domain "machine learning"
| $ wiki write wiki/concepts/attention.md <<'EOF' ... JSON ... EOF
| $ wiki delete wiki/concepts/old.md
| $ wiki search "scaling laws"
| $ wiki lint
|
v
wiki CLI (StorageProvider → local markdown files)
Key principle: The CLI never calls any LLM API. It reads and writes markdown on disk only (no built-in Git sync or cloud backends).
AI assistants / coding agents: Use CLAUDE.md for instructions, rules, and technical context. This README stays oriented to people (overview, install, usage).
Install
npm install -g llmwiki-cli
This gives you two commands: wiki (primary, 4 chars) and llmwiki (fallback if wiki conflicts).
Quick Start
# Create a new wiki
wiki init my-wiki --name "My Notes" --domain "research"
# Write a page (JSON on stdin → YAML frontmatter + body; index updated automatically)
wiki write wiki/concepts/attention.md <<'EOF'
{
"title": "Attention Mechanism",
"tags": ["transformers", "NLP"],
"content": "The attention mechanism allows models to focus on relevant parts of the input.\nSee also [[transformers]] and [[self-attention]]."
}
EOF
# Search and lint
wiki search "attention"
wiki lint
Wiki Structure
When you run wiki init, it creates:
my-wiki/
├── .llmwiki.yaml # Wiki config
├── SCHEMA.md # Instructions for LLM agents
├── raw/ # Immutable source documents
│ └── assets/ # Downloaded images
└── wiki/ # LLM-generated pages
├── index.md # Master index (updated by wiki write / delete)
├── entities/ # People, orgs, products
├── concepts/ # Ideas, frameworks, theories
├── sources/ # One summary per ingested source
└── synthesis/ # Cross-cutting analysis
Use normal Git in my-wiki/ if you want version control. The CLI does not run git init for you.
All markdown pages are stored directly under the wiki root (no profiles/<slug>/ indirection).
Commands
Wiki Management
wiki init [dir] --name <name> --domain <domain> # Create wiki (local files only)
wiki registry # List all wikis
wiki use [wiki-id] # Set active wiki
Reading & Writing
wiki read <path> # Print page markdown to stdout
wiki write <path> # JSON on stdin → frontmatter + body; upserts wiki/index.md for wiki/* paths
wiki delete <path> # Delete page + remove from index
wiki list [dir] [--tree] [--json] # List pages
wiki search <query> [--limit N] [--all] [--json] # Search pages
Health & Links
wiki lint [--json] # Health check
wiki links <path> # Outbound + inbound links
wiki backlinks <path> # Inbound links only
wiki orphans # Pages with no inbound links
wiki status [--json] # Wiki overview stats
LLM Agent Skill Guide
Run wiki skill to print the full guide. The canonical text lives in src/commands/skill.ts (SKILL_GUIDE); update it when commands change. Covers workflows, command patterns, page format, and common gotchas for LLM agents.
How LLM Agents Use This
The generated SCHEMA.md in each wiki contains complete instructions. Here are the typical workflows:
Ingest a Source
# Save raw source (JSON body — large string in "content")
wiki write raw/paper.md <<'EOF'
{"title":"Paper — full text","content":"<paste full text of paper>"}
EOF
# Create structured summary (index line uses title)
wiki write wiki/sources/attention-paper.md <<'EOF'
{
"title": "Attention Is All You Need",
"tags": ["transformers", "attention", "NLP"],
"source": "https://arxiv.org/abs/1706.03762",
"content": "Summary of the attention paper...\nLinks to [[transformers]] and [[self-attention]]."
}
EOF
Answer a Question
wiki search "attention mechanism"
wiki read wiki/concepts/attention.md
wiki links wiki/concepts/attention.md # see related pages
Maintain Wiki Health
wiki lint # find broken links, orphans, missing frontmatter
wiki orphans # pages nobody links to
wiki status # overview stats
Optional: link graph on GitHub Pages
The CLI does not scaffold Git or Actions during wiki init. If you want the same interactive d3-force graph as the live demo:
- Initialize or clone a git repository at your wiki root (
git init, add remote, etc.). - Copy from this repository into your wiki root:
.github/workflows/wiki-viz.yml— source string:getVizWorkflow()insrc/lib/templates.tsscripts/build-graph.jsandscripts/build-site.js—getBuildGraphScript()/getBuildSiteScript()in the same file
Maintainers can regenerate standalone copies withbun scripts/generate-viz-scripts.ts [outDir](writesbuild-graph.cjs/build-site.cjs; rename to.jsif you prefer).
- Commit and push. In the GitHub repo, enable Pages with source GitHub Actions.
The workflow parses [[wikilinks]] on each push and publishes the graph site.
Multi-Wiki Support
The CLI supports multiple wikis via a global registry at ~/.config/llmwiki/:
wiki init ~/wikis/ml --name ml --domain "machine learning"
wiki init ~/wikis/personal --name personal --domain "personal notes"
wiki registry # lists both
wiki use ml # switch active wiki
wiki --wiki personal read wiki/index.md # target specific wiki
wiki search "neural networks" --all # search across all wikis
Wiki resolution order: --wiki flag > cwd .llmwiki.yaml > walk up directories > registry default.
Requirements
- Node.js >= 18 (or Bun)
Development
git clone https://github.com/doum1004/llmwiki-cli
cd llmwiki-cli
bun install
bun test # run full test suite
bun run build # bundle to dist/index.js
bun run dev -- --help
Star History
License
MIT