LLM Wiki Manager
LLM wiki kit
Created by sametbrr
Published in sametbrr/llm-wiki-manager. Publisher page
73 GitHub stars
What it is
Agent Skills plugin for a persistent LLM-managed markdown wiki — ingest, cross-reference, cited queries and lint with 5 stdlib scripts
AutoGineer overview, from the brain’s own files
LLM Wiki Manager is an agent skill and plugin designed to implement the Karpathy LLM Wiki pattern. It shifts the paradigm from traditional RAG—where an LLM re-reads raw documents on every query—to a persistent, compounding knowledge base. The system maintains a clear division of labor: the user curates raw source files in a dedicated folder, while the agent manages the wiki layer, handling synthesis, cross-referencing, indexing, and bookkeeping. It includes idempotent Python scripts for maintenance tasks like linting, logging, and schema evolution, ensuring that knowledge remains structured and traceable over time.
How it works
Memory model: LLM wiki
AutoGineer overview, from the brain’s own files
The system organizes data into a three-layer model: a schema file for conventions, a raw folder for immutable source documents, and a wiki folder for synthesized content. The agent uses a set of Python scripts to manage this structure. When a user provides a source, the agent ingests it, summarizes the content, updates entity and concept pages, and logs the action. It supports advanced operations like semantic updates to propagate changes across multiple pages, automated health checks that generate dated reports, and multi-wiki routing to bridge project-specific and global knowledge bases.
| Folder | Role | Files |
|---|---|---|
| .github | agent-config | 2 |
| skills | agent-instructions | 27 |
| tests | tests | 1 |
| .claude-plugin | folder | 1 |
| .codex-plugin | folder | 1 |
The agent reads
- CLAUDE.md evidence: README.md
The agent writes
- CHANGELOG.md evidence: AGENTS.md
- AGENTS.md evidence: README.md
- CLAUDE.md evidence: README.md
Requirements
Works with: Claude Code, OpenAI Codex, GitHub Copilot, Cursor, Kiro, OpenCode, Amp, Cline, JetBrains Junie, Devin (ex-Windsurf)
Set it up
- Download the brain as a ZIP and unzip it where you keep your notes or project: https://mp-staging.autogineer.com/agent-brains/sametbrr/llm-wiki-manager/download
Claude Code reads CLAUDE.md from this folder automatically.
cd llm-wiki-manager && claude
AutoGineer scanned these files at commit 28ad285.
Created by sametbrr. Published in sametbrr/llm-wiki-manager. Licence: MIT License.
Usage
AutoGineer overview, from the brain’s own files
- Bootstrap a new wiki environment in a research folder.
- Ingest PDF or text documents into the wiki.
- Query the wiki for synthesized answers with citations.
- Perform semantic updates to propagate information changes across the wiki.
- Run automated health checks to identify broken links or orphan pages.
- Route knowledge between project-specific and global wikis.
- “Set up an LLM wiki here. Topic: history of nutrition science.”
- The agent scaffolds the raw/ and wiki/ folders and initializes the schema file.
- “Add this PDF to the wiki.”
- The agent reads the source, writes a summary, updates relevant entity pages, and logs the ingestion.
- “Smith 2024 supersedes Keys 1980, update the wiki.”
- The agent performs a semantic sweep across all pages to update the superseded claim and logs the change.
- “Lint the wiki.”
- The agent runs a health check, generates a dated report in the reports folder, and updates the index.
Documentation by sametbrr
Image: Release Image: License: MIT Image: Agent Skills
Quick Start • Installation • Usage • Update mode • Multi-wiki • Compatibility
LLM Wiki Manager
An Agent Skill and plugin for Claude Code and Codex for building and maintaining a personal LLM-managed wiki — a persistent, compounding knowledge base where the LLM does all the writing, cross-referencing, and bookkeeping while you curate sources and ask questions.
🇹🇷 Türkçe için README.tr.md
Quick Start
claude plugin marketplace add sametbrr/skill-hub && claude plugin install llm-wiki-manager@skill-hub # Claude Code
codex plugin marketplace add sametbrr/skill-hub && codex plugin add llm-wiki-manager@skill-hub # CodexStart a new session in your research folder:
mkdir ~/research/my-topic && cd ~/research/my-topic && claude # or: codex
> "Set up an LLM wiki here. Topic: history of nutrition science."Features
Instead of RAG — where the LLM rediscovers answers from raw documents on every query — this pattern has the LLM compile raw sources into a persistent, interlinked markdown wiki. Each new source enriches existing pages. Cross-references are built eagerly. Contradictions are flagged. Knowledge compounds over time.
Implements Karpathy's LLM Wiki pattern as a full agent skill with 9 operating modes (including multi-wiki routing and migrate), 5 idempotent Python scripts, 8 page templates, and 11 reference documents.
Without this pattern With this pattern
──────────────────── ──────────────────
Query 1 → re-read 50 docs Query 1 → read compiled wiki (already synthesized)
Query 2 → re-read 50 docs Query 2 → read updated wiki (cross-refs already there)
Query 3 → re-read 50 docs Query 3 → read updated wiki (contradictions already flagged)Requirements
Claude Code, OpenAI Codex, or any agentskills.io-compatible agent
Python 3.9+ (stdlib only, for the 5 included scripts — no pip install needed)
Installation
Option 1 — Claude Code plugin (recommended)
claude plugin marketplace add sametbrr/skill-hub
claude plugin install llm-wiki-manager@skill-hubOption 2 — Codex plugin
codex plugin marketplace add sametbrr/skill-hub
codex plugin add llm-wiki-manager@skill-hubOption 3 — GitHub CLI, any agent (gh CLI with gh skill)
gh skill install sametbrr/llm-wiki-manager llm-wiki-manager --agent claude-code --scope user # or --agent codexOption 4 — manual copy: copy skills/llm-wiki-manager/ into ~/.claude/skills/ or ~/.agents/skills/.
After installing, start a new session. The skill loads automatically when relevant.
Usage
The skill auto-detects which mode applies from natural language. No slash commands needed.
> "Set up an LLM wiki here" # Bootstrap: scaffold raw/, wiki/ and the schema file
> "Add this PDF to the wiki" # Ingest: summarize, cross-reference, index, log
> "What does the wiki say about X?" # Query: answer with citations, offer to file back
> "Smith 2024 supersedes Keys 1980, update the wiki" # Update: sweep every page that paraphrases the claim
> "Lint the wiki" # Lint: dated health report in wiki/reports/
> "We should always do X going forward" # Schema-evolve: record a convention for future sessions
> "Add this to my global wiki" # Multi-wiki: route between project and global wiki
> "How does this pattern work?" # Teach: explain the pattern and compare it with RAGModes
| Mode | Trigger examples | What happens |
|---|---|---|
| Bootstrap | "Set up a wiki", "start a knowledge base here" | Scaffolds raw/, wiki/ and the schema file (AGENTS.md + a CLAUDE.md pointer) from templates |
| Ingest | "Add this PDF to the wiki", "I just read X, file it" | Reads source → writes summary → updates entity/concept pages → indexes → logs |
| Query | "What does the wiki say about X?", "Compare X and Y" | Reads index → candidate pages → synthesizes answer with citations → offers to file back |
| Update | "Smith 2024 supersedes Keys 1980, update the wiki" | Semantic sweep across all pages → diff-before-write per page → single log entry |
| Lint | "Health check the wiki", "anything broken?" | Runs lint_wiki.py → auto-saves wiki/reports/lint-YYYY-MM-DD.md → auto-tracks in index and log |
| Schema-evolve | "We should always do X going forward" | Updates the schema file (AGENTS.md / CLAUDE.md) so future sessions inherit the convention |
| Multi-wiki | "Add this to my global wiki", "promote this page to global" | Routes between project wiki and global wiki using the External Wiki: declaration in project CLAUDE.md |
| Teach | "How does this pattern work?", "explain the LLM wiki idea" | Explains the pattern, compares with RAG, walks through a concrete example |
Full walkthrough
# 1. Go to your research folder
mkdir ~/research/my-topic && cd ~/research/my-topic && claude
# 2. Bootstrap the wiki
> "Set up an LLM wiki here. Topic: history of nutrition science."
# 3. Drop a source
cp ~/Downloads/pollan-2008.pdf raw/
# 4. Ingest it
> "Ingest Pollan's In Defense of Food"
# 5. Ask questions
> "What does the wiki say about nutritionism?"
# 6. Health check (auto-saves dated report to wiki/reports/)
> "Lint the wiki"The three-layer model
your-wiki/
├── AGENTS.md # Schema — conventions for this wiki (co-evolved over time)
├── CLAUDE.md # One line: @AGENTS.md (Claude Code reads CLAUDE.md, Codex reads AGENTS.md)
├── raw/ # YOUR layer — immutable sources you curate. LLM reads, never writes.
└── wiki/ # LLM layer — all pages written and maintained by the LLM
├── index.md # Content catalog (updated on every ingest)
├── log.md # Append-only operation log (greppable)
├── hot.md # Hot cache — most recently ingested sources and active references
├── sources/ # One summary page per ingested source
├── entities/ # People, organizations, places, products
├── concepts/ # Ideas, theories, frameworks, terms
├── notes/ # Filed-back query answers and loose pages
└── reports/ # Auto-generated dated lint reports (wiki/reports/lint-YYYY-MM-DD.md)Division of labor:
| You do | The LLM does |
|---|---|
| Curate sources (decide what to read) | Read sources end-to-end |
| Ask questions, steer direction | Write summaries, entity and concept pages |
| Review the wiki, follow links | Update cross-references in-place |
| Decide what matters | Maintain index.md and log.md |
Own raw/ | Flag contradictions, surface gaps |
You almost never write wiki pages by hand. The LLM does the bookkeeping — that's what makes the wiki compound instead of collapse.
Core disciplines
LLM owns
wiki/. You ownraw/. No exceptions.Every operation logs to
log.mdviaappend_log.py. Greppable:grep "^## \[" log.md | tail -20Every new or updated page touches
index.mdviaupdate_index.py. Stale index = wiki that feels lost.Cross-reference aggressively. When a source mentions an entity that already has a page, update that page. Don't leave connections implicit.
Cite back to
raw/. Every claim is traceable to a specific source file.Flag contradictions, don't overwrite. New source disagrees with old claim? Both stay, marked with their source, with a
> [!warning] Sources disagreecallout.Schema lives in
CLAUDE.md. When a convention works, write it down. The next session starts informed.
What's inside
Scripts (Python stdlib, no dependencies, all idempotent)
| Script | Purpose |
|---|---|
scripts/init_wiki.py | Scaffold a new wiki — creates raw/, wiki/, the schema file (AGENTS.md + CLAUDE.md pointer, or keeps an existing one), index.md, log.md, and hot.md. Idempotent. |
scripts/append_log.py | Append a ## [YYYY-MM-DD] action | title entry to log.md. Supports flexible log path detection. |
scripts/update_index.py | Add or update an entry under a category in index.md. Case-insensitive categories, keeps section order and content, follows the index link style (markdown or [[wikilinks]]), new categories go before ## Meta. |
scripts/lint_wiki.py | Health check. Detects orphan pages, broken links and index drift in both standard markdown and Obsidian wiki-link format ([[page#heading]], ![[embed.png]] included). Default: writes wiki/reports/lint-<today>.md and auto-tracks. Run --stdout for terminal output. |
scripts/migrate_wiki.py | Schema upgrade (v1 → v2). Deduplicates index.md, moves dated changelog blocks from hot.md into log.md, stamps the schema version. Idempotent. |
Templates
| Template | Used for |
|---|---|
wiki-CLAUDE.md.tmpl | The schema file dropped into a fresh wiki |
source-summary.md.tmpl | One ingested source — claims, methodology, cross-links, open questions |
entity-page.md.tmpl | People, organizations, places, products |
concept-page.md.tmpl | Ideas, frameworks, theories, terms |
comparison-page.md.tmpl | "X vs Y" pages, often filed-back query answers |
index.md.tmpl | Initial content catalog |
log.md.tmpl | Initial log with bootstrap entry |
hot.md.tmpl | Initial hot cache — rewritten after each ingest with latest sources and active references |
Reference docs
Eleven detailed workflow documents in references/:
philosophy.md · architecture.md · bootstrap-workflow.md · ingest-workflow.md · query-workflow.md · update-workflow.md · lint-workflow.md · migrate-workflow.md · schema-design-guide.md · multi-wiki-routing.md · teaching-mode.md
The skill reads these selectively — you don't need to. They're there to give the LLM depth on each mode.
Update mode
Standard ingest already handles contradictions on a single page (Disputes section). Update mode is for when a new source supersedes a claim that's paraphrased across multiple pages — the same idea written four different ways in four different files.
Scenario: Smith 2024 reanalysis shows Keys 1980's seven-countries study cherry-picked data.
The Keys r=0.87 claim appears as:
concepts/saturated-fat.md → "Keys found r=0.87 across seven countries"
entities/ancel-keys.md → "famous for showing strong correlation"
concepts/heart-disease.md → "saturated fat is a primary driver per Keys"
concepts/dietary-policy.md → "the saturated fat hypothesis drove decades of policy"
Update mode:
1. Semantic sweep — finds all four (grep won't, LLM will)
2. Shows scope: "4 pages affected. Proceed?"
3. Diff-before-write per page — y/n/skip/edit per change
4. Per-page strategy: revise / disputes / annotate (not one-size-fits-all)
5. One log entry tying all four edits to Smith 2024This is change propagation, not change tracking. The log and frontmatter track what was edited. Update mode ensures the edit reaches everywhere it should.
Auto-dated lint reports
lint_wiki.py with no flags:
Writes
wiki/reports/lint-YYYY-MM-DD.md(overwrites same-day run — idempotent daily)Adds a
Reportsindex entry automaticallyAppends a
lint | Health checklog entry automaticallyExits with code 1 if any block-severity issue found (useful for CI)
Reports accumulate as a longitudinal health record. git log wiki/reports/ shows how wiki quality evolved over time.
Override flags: --stdout (terminal, no tracking), --no-track (write file, skip index/log), --report PATH (custom path).
Multi-wiki
Most users start with one wiki. Once you have two — say, a per-project wiki at the working directory plus a long-lived global "second brain" (often an existing Obsidian vault) — the skill routes writes between them based on a single declaration in the project's CLAUDE.md.
~/projects/x-project/ ← active project (current working directory)
├── CLAUDE.md ← project schema — declares the global wiki path
├── raw/ ← project sources
└── wiki/ ← project wiki
~/Documents/obsidian/ ← global wiki (long-lived, exists across projects)
├── CLAUDE.md ← global schema
├── raw/
└── wiki/One-time setup
Add this to the project's CLAUDE.md (or ask the agent to do it):
## External Wiki
Global knowledge base: ~/Documents/obsidian/
### Routing rules
- Project-specific code decisions, architecture, bugs, configuration → this project's `wiki/`
- Concepts, frameworks, patterns, ideas applicable beyond this project → global wiki
- When in doubt, ask before writing
- Scripts always need `--path` flag pointing to the right wiki root
### Cross-wiki links
- Use absolute paths (`~/...`) when linking from one wiki to the other.
- Never use relative paths that cross wiki boundaries.Four canonical scenarios
| # | Scenario | Trigger | What the agent does |
|---|---|---|---|
| A | Write to global while in a project | "Add JWT refresh rotation to my global wiki" | Reads project CLAUDE.md → resolves global path → writes to global, project wiki untouched |
| B | Pull from global into a project | "What does the global wiki say about rate limiting? Apply it to /api/search" | Reads global pages → synthesizes recommendation → writes a project-specific page that links to the global one (never copies) |
| C | Promote a page from project to global | "concepts/event-sourcing.md has matured, promote it" | Moves content to global → leaves a one-line redirect stub at the project path → updates both indexes and logs |
| D | Lint both wikis at once | "Lint both wikis" | Runs lint_wiki.py --path against each → reads both reports → returns one summary |
Full walkthroughs of all four scenarios in references/multi-wiki-routing.md.
Compatibility
| Tool | Install | Schema file it reads |
|---|---|---|
| Claude Code | plugin (claude plugin install llm-wiki-manager@skill-hub) or ~/.claude/skills/ | CLAUDE.md |
| OpenAI Codex | plugin (codex plugin add llm-wiki-manager@skill-hub) or ~/.agents/skills/ | AGENTS.md |
| GitHub Copilot, Cursor, Gemini CLI, … | gh skill install … --agent <name> | usually AGENTS.md |
A wiki that keeps its schema in AGENTS.md with a one-line CLAUDE.md (@AGENTS.md) works in all of them; init_wiki.py sets that up.
Related
Karpathy's LLM Wiki gist — the original idea
agentskills.io — open standard spec
Contributing
Questions and ideas: Discussions
Bugs: issue tracker
Pull requests: read the contribution guide first
Security reports: follow the security policy, not a public issue
License
MIT — see LICENSE.
<div align="center">
</div>
skills/llm-wiki-manager/references/ingest-workflow.md
Ingest Workflow
The most-used mode. Use this when the user drops a new file in raw/, pastes a URL or quote, attaches a PDF, or says something like "I just read X, let's add it" or "file this".
This is where the wiki actually compounds. Done well, every ingest leaves the wiki noticeably richer. Done lazily — just a flat summary, no cross-references — and the wiki is just RAG with extra steps.
Goal
Read the source carefully, integrate its information into the existing wiki at every relevant point, and leave a clear trail (source-summary page, updated entity/concept pages, index entries, log entry). The user should be able to come back in three months and see exactly what was added and why.
The full loop
1. Locate and read the source
If the user dropped a file: confirm the path under raw/. If they pasted a URL: download it (use whatever fetch tool is available), save the markdown/HTML/PDF to raw/ with a sortable filename (YYYY-MM-author-short-title.ext), then read it. If they pasted text directly: write it to raw/ first — never ingest something that doesn't have a stable file to cite back to.
Read the source end-to-end before writing anything. For long sources (books, long papers), read in passes:
First pass — skim for structure, key claims, named entities, key concepts.
Second pass — re-read the parts that matter for the wiki.
For PDFs with images that matter (charts, diagrams), view the relevant pages as images. For audio/video transcripts, read the transcript and note timestamps for the standout moments.
2. Survey the existing wiki
Before writing, read wiki/index.md to understand what's already there. Then read pages that look related — entities you recognize, concepts that overlap. This is what makes the ingest integrative rather than additive.
A common failure: writing a source summary that mentions "Pollan" without checking that wiki/entities/michael-pollan.md already exists. The integration disappears. Always check first.
3. Discuss with the user (briefly)
Before producing pages, give a 4-6 line summary of what you saw and what you plan to do:
Read In Defense of Food. Key takeaways: (1) "Eat food, not too much, mostly plants" — Pollan's full thesis; (2) coins nutritionism as critique of reductive nutrition science; (3) heavy critique of the USDA food pyramid era. Existing pages I'd update:
entities/michael-pollan.md(extend),concepts/nutritionism.md(new). Want me to proceed, or anything to emphasize / leave out?
This 30-second checkpoint catches misdirected ingests early. The user might say "skip the USDA stuff, that's not relevant to this wiki" or "actually focus on the supplements chapter". Cheap to ask, expensive to redo.
If the user has set up the wiki for batch-ingest with low supervision, skip this step. Note it in CLAUDE.md if so.
4. Write the source summary
Create wiki/sources/<slug>.md using assets/templates/source-summary.md.tmpl. Standard sections:
Frontmatter (optional): type, date, length, status (read/skim/excerpt)
One-line summary at the top
Key claims as a bulleted list, each with a source reference (page number, timestamp, paragraph)
Methodology / evidence quality — useful especially for papers, studies, journalism
How this relates to the wiki — explicit pointers to entity/concept pages this source touches
Notable quotes — sparingly, with citation
Open questions — things this source raised that aren't answered
The summary is not a substitute for the source. Don't try to compress everything; capture what's load-bearing for future synthesis.
5. Update entity and concept pages
For each entity or concept the source touches significantly:
If the page exists: read it, then update in place. Add a new bullet under the relevant section, extend an existing claim, or add a "Disputes" subsection if the source contradicts something already there. Use the source-summary page as a cite.
If the page doesn't exist but the entity/concept is substantial enough to warrant one: create it from
assets/templates/entity-page.md.tmplorconcept-page.md.tmpl. Don't create pages for entities that are only mentioned in passing — those can stay in the source summary.
A rule of thumb: an entity or concept earns a page once it's mentioned in two or more sources, or it's central to the user's purpose. Until then, keep it in source summaries.
6. Cross-link
Link from each updated page back to the source summary, and from the source summary to the entity/concept pages. The connectivity is the value.
Standard markdown link style:
See [Nutritionism](../concepts/nutritionism.md) for Pollan's coined term.
This contradicts [USDA's 1992 pyramid](../entities/usda.md#1992-food-pyramid).If the wiki is set up for Obsidian-style [[wikilinks]], use those — check CLAUDE.md.
Maintain the ## Related footer. Every page ends with a ## Related section: 2–5 links, each with a one-line why (- [Page](path.md) — same program, parallel track). When this ingest creates a meaningful connection between two pages, add it to both pages' Related footers. Don't mirror the links into a related: frontmatter field — the footer is the single source of truth.
Hub bookkeeping. If a topic cluster reaches 3+ pages with this ingest, elect its most encompassing page as the hub: add a ## Pages in this cluster section there, and mark it with ★ in the index (see CLAUDE.md hub rules).
7. Flag contradictions
If the new source contradicts a claim on an existing page, don't overwrite. Add the new claim alongside the old one, mark both with their source, and note the conflict:
## Effect of saturated fat on heart disease
Two sources disagree:
- [Keys 1980](../sources/keys-1980.md): saturated fat strongly correlates with heart disease incidence (r=0.87 across seven countries).
- [Teicholz 2014](../sources/teicholz-big-fat-surprise.md): the seven-country selection cherry-picked outcomes; broader sample shows weak correlation.
> [!warning] Sources disagree
> Open question for further investigation.This is the "contradictions are flagged" part of the value proposition. Silent overwrites destroy that.
When the contradiction spans multiple pages, the local Disputes pattern is not enough. If you notice that the new source's claim invalidates content on 3 or more existing pages (entities, other concepts, notes), stop and switch to the update workflow (references/update-workflow.md). It runs a semantic sweep across the wiki, shows diffs page-by-page, and propagates the correction with one coherent log entry. Use it when the same idea is paraphrased across the wiki and patching one page leaves the others stale.
8. Update the index
Run scripts/update_index.py for each new page and for any page whose summary changed substantially:
python scripts/update_index.py --category concepts \
--title "Nutritionism" \
--page-path "wiki/concepts/nutritionism.md" \
--summary "Reductive ideology that food = nutrients (5 sources)"The script is idempotent on (category, title) — re-running updates the existing entry rather than duplicating.
9. Append to log
One entry. Use scripts/append_log.py:
python scripts/append_log.py --action ingest \
--title "In Defense of Food (Pollan, 2008)" \
--details "Created sources/pollan-eat-food.md. New: concepts/nutritionism.md. Updated: entities/michael-pollan.md, entities/usda.md. Touched 4 pages."The log entry should be specific enough that "what did I do last week" returns useful answers later.
10. Rewrite hot.md
If the wiki has a hot.md (at wiki/hot.md or at the wiki root), rewrite it entirely — do not append. The file has exactly five sections (see assets/templates/hot.md.tmpl), ~500 words total:
State: one paragraph — page count, layout, active policies.
Last ingest: ONLY the most recent 1-2 operations; everything older lives in log.md. Never accumulate dated
## [...]blocks here — those are log entries.Active themes: one line per theme + link to its hub (
★) page.Open items: add anything surfaced during ingest (a gap in coverage, an open question, a follow-up source worth finding); remove resolved ones.
Conventions: 3-4 lines a cross-project reader must know; details live in CLAUDE.md.
If this ingest introduced a new tag, add it to the canonical tag list in CLAUDE.md (if it fits the tag policy) — the tag inventory does NOT live in hot.md.
Rewriting (not appending) is important: hot.md is a cache, not a log. Its job is to give a future session — or a cross-project reader — an instant orientation. An append-only hot.md becomes too long to serve that purpose; lint_wiki.py flags it when it bloats.
11. Brief recap to the user
End with a short summary: pages created, pages updated, anything notable (a strong contradiction with existing material, an open question worth investigating, a missing source that would help). Three to six lines.
If the user is engaged and curious, offer follow-ups: "This source gestures at [some related thing] but doesn't develop it. Want to find a source that does?"
Heuristics
Five to fifteen pages touched is normal for a substantive source. Two pages touched usually means the integration was lazy. Twenty-five pages touched usually means the source got over-integrated and trivial mentions got their own pages.
Source summary first, then ripple outward. Write the source summary, then update pages it points to, then update the index, then log. Don't bounce around.
Quote sparingly. A wiki page is a synthesis, not a copy. Direct quotes are reserved for moments where the exact wording matters.
Match the schema. Read
CLAUDE.mdif there's any doubt about page structure, frontmatter, or category names. The schema is the authority.
Common mistakes
Skipping the survey step. Writing as if the wiki is empty when it isn't. Cross-references die quietly.
One giant page per source. No, the source summary is one page; the entities and concepts spread across many pages. The summary points outward, not inward.
Creating premature entity pages. A name mentioned in passing once is not an entity yet. Wait until it earns the page.
Forgetting the index or the log. Both are easy to skip and both quietly degrade the wiki. The scripts are short; just run them.
Silent contradiction handling. Overwriting old claims with new ones. Flag instead.
Duplicating content across pages. If a fact is needed in two pages, write it once and link.
40 files
- .claude-plugin/plugin.json
- .codex-plugin/plugin.json
- .github/dependabot.yml
- .github/workflows/release.yml
- .gitignore
- AGENTS.md
- ATTRIBUTION.md
- CHANGELOG.md
- CLAUDE.md
- LICENSE
- README.md
- README.tr.md
- skills/llm-wiki-manager/SKILL.md
- skills/llm-wiki-manager/assets/examples/healthy-wiki-tree.txt
- skills/llm-wiki-manager/assets/templates/comparison-page.md.tmpl
- skills/llm-wiki-manager/assets/templates/concept-page.md.tmpl
- skills/llm-wiki-manager/assets/templates/entity-page.md.tmpl
- skills/llm-wiki-manager/assets/templates/hot.md.tmpl
- skills/llm-wiki-manager/assets/templates/index.md.tmpl
- skills/llm-wiki-manager/assets/templates/log.md.tmpl
- skills/llm-wiki-manager/assets/templates/source-summary.md.tmpl
- skills/llm-wiki-manager/assets/templates/wiki-CLAUDE.md.tmpl
- skills/llm-wiki-manager/references/architecture.md
- skills/llm-wiki-manager/references/bootstrap-workflow.md
- skills/llm-wiki-manager/references/ingest-workflow.md
- skills/llm-wiki-manager/references/lint-workflow.md
- skills/llm-wiki-manager/references/migrate-workflow.md
- skills/llm-wiki-manager/references/multi-wiki-routing.md
- skills/llm-wiki-manager/references/philosophy.md
- skills/llm-wiki-manager/references/query-workflow.md
- skills/llm-wiki-manager/references/schema-design-guide.md
- skills/llm-wiki-manager/references/teaching-mode.md
- skills/llm-wiki-manager/references/update-workflow.md
- skills/llm-wiki-manager/scripts/append_log.py
- skills/llm-wiki-manager/scripts/init_wiki.py
- skills/llm-wiki-manager/scripts/lint_wiki.py
- skills/llm-wiki-manager/scripts/migrate_wiki.py
- skills/llm-wiki-manager/scripts/schema_file.py
- skills/llm-wiki-manager/scripts/update_index.py
- tests/test_scripts.py
Advantages
- Compounding knowledge base that synthesizes information rather than re-reading raw documents on every query. evidence: README.md
- Automated bookkeeping and cross-referencing reduce manual maintenance effort. evidence: README.md
- Idempotent scripts ensure reliable operations without external dependencies. evidence: README.md
Limitations
- Requires strict adherence to the division of labor between the raw source folder and the LLM-managed wiki folder. evidence: README.md
- Multi-wiki routing requires manual configuration of paths in the schema file. evidence: README.md
Security
Risk level: No risk found
This brain runs
Agent instructions that run or fetch something (8)
AGENTS.md:`tests/test_scripts.py`: functional tests that run every script against temp wikis.AGENTS.md:`python3 -m unittest discover -s tests`AGENTS.md:Codex: install from a local marketplace in an isolated `CODEX_HOME`;skills/llm-wiki-manager/SKILL.md:Run them by path relative to this skill's directory, and always pass `--path <wiki-root>`.skills/llm-wiki-manager/SKILL.md:If it is lower than the version this skill expects, offer one sentence before starting the requested work: *"This wiki is at schema v1, the skill expects v2 — run `scripts/migrate_wiki.py` first?"* Don't block the requested operation if the user declines;skills/llm-wiki-manager/SKILL.md:If the section is missing and a multi-wiki operation is requested, **run Schema-evolve first** to add it, then proceed.skills/llm-wiki-manager/SKILL.md:Some users run two wikis: a per-project wiki at the working directory and a long-lived global wiki (often an existing Obsidian vault) at a fixed path.skills/llm-wiki-manager/SKILL.md:unstamped = v1), shows a **dry-run** of the mechanical steps plus the manual (LLM) steps, applies with `--apply` only.
Attribution and licence
- Created by
- sametbrr
- Declared authors
- sametbrr (declared in .claude-plugin/plugin.json)
- Publisher
- Samet Birer (Community)
- Source
- sametbrr/llm-wiki-manager · commit 28ad285
- Licence
- MIT License (files served here)
- ATTRIBUTION.md (written by AutoGineer)
- Synced
AutoGineer is not affiliated with or endorsed by sametbrr.
Download
The ZIP holds the files AutoGineer serves, with ATTRIBUTION.md and the licence. Download llm-wiki-manager.zip