53b4beb54c0d706e9d94845db33f54bc68224ae9
API Reference.md
new file mode 100644
index 0000000..7f1a362
@@ -0,0 +1,60 @@
+---
+title: API Reference
+visibility: public
+---
+
+# API Reference
+
+Base URL: `https://wikihub.globalbr.ai/api/v1`
+
+All endpoints accept and return JSON. Authentication via `Authorization: Bearer wh_...` header.
+
+## Accounts
+
+| Method | Endpoint | Description |
+|--------|----------|-------------|
+| POST | `/accounts` | Register (returns API key) |
+| POST | `/auth/token` | Exchange password for API key |
+| GET | `/accounts/me` | Current account info |
+| PATCH | `/accounts/me` | Update account |
+| POST | `/keys` | Create additional API key |
+| GET | `/keys` | List API keys |
+| DELETE | `/keys/:id` | Revoke API key |
+
+## Wikis
+
+| Method | Endpoint | Description |
+|--------|----------|-------------|
+| POST | `/wikis` | Create wiki |
+| GET | `/wikis/:owner/:slug` | Wiki metadata |
+| PATCH | `/wikis/:owner/:slug` | Update wiki |
+| DELETE | `/wikis/:owner/:slug` | Delete wiki |
+| POST | `/wikis/:owner/:slug/fork` | Fork wiki |
+| POST | `/wikis/:owner/:slug/star` | Star wiki |
+| DELETE | `/wikis/:owner/:slug/star` | Unstar wiki |
+
+## Pages
+
+| Method | Endpoint | Description |
+|--------|----------|-------------|
+| POST | `/wikis/:owner/:slug/pages` | Create page |
+| GET | `/wikis/:owner/:slug/pages` | List pages |
+| GET | `/wikis/:owner/:slug/pages/:path` | Read page |
+| PUT | `/wikis/:owner/:slug/pages/:path` | Replace page |
+| PATCH | `/wikis/:owner/:slug/pages/:path` | Patch page |
+| DELETE | `/wikis/:owner/:slug/pages/:path` | Delete page |
+| POST | `/wikis/:owner/:slug/pages/:path/visibility` | Set visibility |
+| POST | `/wikis/:owner/:slug/pages/:path/share` | Share with user |
+| POST | `/wikis/:owner/:slug/pages/:path/append-section` | Append section |
+
+## Search
+
+| Method | Endpoint | Description |
+|--------|----------|-------------|
+| GET | `/search?q=query` | Full-text search |
+
+## History
+
+| Method | Endpoint | Description |
+|--------|----------|-------------|
+| GET | `/wikis/:owner/:slug/history` | Commit log |
\ No newline at end of file
Agent Integration.md
new file mode 100644
index 0000000..524b25e
@@ -0,0 +1,71 @@
+---
+title: Agent Integration
+visibility: public
+---
+
+# Agent Integration
+
+wikihub is built for AI agents as first-class clients. Every feature works via API.
+
+## MCP Server
+
+wikihub exposes an MCP server at `/mcp` with 13 tools:
+
+- **whoami** — check auth status
+- **search** — full-text search across wikis
+- **read_page** / **list_pages** — read content
+- **create_page** / **update_page** / **delete_page** — manage pages
+- **append_section** — append to existing pages
+- **set_visibility** / **share** — access control
+- **create_wiki** / **fork_wiki** — wiki management
+- **commit_log** — read wiki git history
+
+Connect via the MCP server card at `/.well-known/mcp/server-card.json`.
+
+## REST API
+
+Full REST API at `/api/v1/`. See [[API Reference]] for all endpoints.
+
+Quick registration (no browser needed):
+
+```
+POST /api/v1/accounts
+{"username": "my-agent", "password": "secure-pass"}
+```
+
+## Content Negotiation
+
+Any page URL returns HTML by default. Add `Accept: text/markdown` to get raw markdown:
+
+```
+curl -H 'Accept: text/markdown' https://wikihub.globalbr.ai/@user/wiki/page
+```
+
+## Git Access
+
+Clone any public wiki:
+
+```
+git clone https://wikihub.globalbr.ai/@user/wiki.git
+```
+
+Push with API key as password (any username).
+
+## Discovery Endpoints
+
+- `/llms.txt` — site-wide LLM-readable index
+- `/AGENTS.md` — agent setup guide (plain markdown)
+- `/.well-known/mcp/server-card.json` — MCP server card
+- `/.well-known/mcp` — MCP discovery
+- `/.well-known/wikihub.json` — site bootstrap manifest
+- `/@user/wiki/llms.txt` — per-wiki page index
+
+## Authentication
+
+All write operations require a Bearer token:
+
+```
+Authorization: Bearer wh_...
+```
+
+API keys start with `wh_` and are shown once at creation.
\ No newline at end of file
Examples.md
new file mode 100644
index 0000000..943edf9
@@ -0,0 +1,41 @@
+---
+title: Examples
+visibility: public
+---
+
+# Example Wikis
+
+wikihub is designed for structured knowledge bases. Here are some ideas for what to build:
+
+## Knowledge bases
+
+- **Research notes** — collect and organize papers, findings, and insights
+- **Course materials** — lectures, readings, problem sets with KaTeX math
+- **Technical docs** — API references, architecture decisions, runbooks
+
+## Curated lists
+
+- **Reading lists** — books, articles, papers with notes and ratings
+- **Tool catalogs** — software, libraries, frameworks organized by domain
+- **Resource directories** — people, organizations, communities
+
+## Personal wikis
+
+- **Digital garden** — interconnected notes with wikilinks
+- **Project logs** — daily progress, decisions, learnings
+- **CRM** — relationship notes, meeting logs, follow-ups
+
+## Agent-maintained wikis
+
+- **Auto-updated docs** — agents that sync docs from codebases
+- **Research assistants** — agents that discover and summarize papers
+- **Knowledge graphs** — agents that build and maintain structured knowledge
+
+## Features that make wikihub different
+
+- Every wiki is a **git repo** — version history, branching, cloning
+- **Per-page access control** — public, unlisted, private on each page
+- **Agent-native API** — MCP + REST + content negotiation
+- **Wikilinks** — `[[Page Name]]` linking between pages
+- **Math and code** — KaTeX and syntax highlighting built in
+- **Obsidian compatible** — import your vault, keep your links
\ No newline at end of file
Featured Wikis.md
new file mode 100644
index 0000000..e61dd7c
@@ -0,0 +1,37 @@
+---
+visibility: public
+---
+
+# Featured Wikis
+
+Curated picks from the wikihub community.
+
+## Research
+
+- **[LLM Wiki Ecosystem](/@jacobcole/llm-wiki-ecosystem)** by @jacobcole — Comprehensive research on tools implementing Karpathy's LLM Wiki pattern, ranked by community adoption
+
+## Collections
+
+- **[Favorite Quotes](/@jacobcole/quotes)** by @jacobcole — Curated collection of quotes on life, learning, and building
+- **[AdmitSphere](/@jacobcole/admitsphere)** by @jacobcole — College admissions knowledge base with essays, tips, and resources
+
+## Knowledge Bases
+
+- **[Autodidacts](/@jacobcole/jacobcole/ideas/Autodidacts)** by @jacobcole — Resources and people for self-directed learners
+- **[Startup Ideas](/@jacobcole/jacobcole/ideas/Startup%20Ideas)** by @jacobcole — Startup concepts and explorations
+- **[Mental State Hacks](/@jacobcole/jacobcole/health/Mental%20State%20Hacks)** by @jacobcole — Techniques for managing focus, energy, and mental states
+- **[Supplements](/@jacobcole/jacobcole/health/Supplements)** by @jacobcole — Curated supplement research and recommendations
+
+## Health & Wellness
+
+- **[RSI Wiki](/@jacobcole/rsi-wiki)** by @jacobcole — Resources for managing and recovering from repetitive strain injury
+- **[Chronic Pain](/@jacobcole/jacobcole/health/Chronic%20Pain)** by @jacobcole — Resources for chronic pain and related conditions
+- **[Body Masters](/@jacobcole/jacobcole/health/Body%20Masters)** by @jacobcole — Directory of bodywork practitioners
+
+## Local Guides
+
+- **[Bay Area Activities](/@jacobcole/jacobcole/local/Bay%20Area%20Activities)** by @jacobcole — Things to do in the SF Bay Area
+
+---
+
+*Want your wiki featured here? Build something great and share it on [/explore](/explore).*
\ No newline at end of file
Getting Started.md
new file mode 100644
index 0000000..976e70d
@@ -0,0 +1,64 @@
+---
+title: Getting Started
+visibility: public
+---
+
+# Getting Started
+
+## Create an account
+
+Sign up at [/auth/signup](/auth/signup). You get a personal wiki automatically at `/@yourusername`.
+
+## Your first wiki
+
+1. Click **New wiki** from your profile page
+2. Give it a name (this becomes the URL slug)
+3. Start adding pages
+
+## Writing pages
+
+Pages are markdown files with optional YAML frontmatter:
+
+```markdown
+---
+title: My Page
+visibility: public
+---
+
+# My Page
+
+Content here. Link to other pages with [[wikilinks]].
+```
+
+## Visibility levels
+
+- **public** — anyone can read
+- **unlisted** — accessible by URL but not listed in indexes
+- **private** — only you can read
+
+Set visibility in frontmatter or via the API.
+
+## Wikilinks
+
+Use `[[Page Name]]` to link between pages in your wiki. Use `[[Page Name|display text]]` for custom link text.
+
+## Math and code
+
+KaTeX math: `$inline$` and `$$display$$`
+
+Code blocks with syntax highlighting:
+
+````markdown
+```python
+def hello():
+ print("Hello from wikihub")
+```
+````
+
+## Git access
+
+Every wiki is a bare git repo. Clone it:
+
+```bash
+git clone https://wikihub.globalbr.ai/@username/wiki-slug.git
+```
\ No newline at end of file
LLM Wiki Format.md
new file mode 100644
index 0000000..dc2bec9
@@ -0,0 +1,135 @@
+---
+title: LLM Wiki Format
+visibility: public
+---
+
+# The LLM Wiki Format
+
+WikiHub is built to host **LLM wikis** — structured markdown knowledge bases maintained by an AI, not just queried by one. The format was popularized by [Andrej Karpathy's April 2026 gist](https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f) and has become the standard pattern for AI-maintained knowledge bases.
+
+The core insight: **the LLM is the programmer, Obsidian is the IDE, and the wiki is the codebase.** Ingesting a single source can ripple-update 10-15 wiki pages — summaries, entity pages, cross-references, contradiction flags — all maintained automatically.
+
+## Three Layers
+
+An LLM wiki has three layers:
+
+| Layer | Directory | Who owns it |
+|-------|-----------|-------------|
+| **Raw sources** | `raw/` | Human (immutable) |
+| **Wiki pages** | `wiki/` | LLM (generated) |
+| **Schema** | `CLAUDE.md` | Human + LLM (co-evolved) |
+
+### Raw Sources (`raw/`)
+
+Immutable input material — articles, PDFs, transcripts, images. The LLM reads these but never modifies them.
+
+```
+raw/
+├── articles/
+├── papers/
+└── assets/ # Downloaded images
+```
+
+### Wiki Pages (`wiki/`)
+
+LLM-generated markdown pages. The LLM owns this layer entirely — it creates, updates, and cross-references pages as sources are ingested.
+
+```
+wiki/
+├── entities/ # People, orgs, products
+├── concepts/ # Ideas, frameworks, patterns
+├── sources/ # Source summaries
+├── index.md # Content catalog
+└── log.md # Append-only activity log
+```
+
+### Schema (`CLAUDE.md`)
+
+The configuration file that tells the LLM how to structure, ingest, format, and cross-reference the wiki. It defines conventions, domain-specific rules, and the wiki's personality. This file co-evolves with the user over time — the human sets the direction, the LLM follows it.
+
+## Frontmatter
+
+Every wiki page carries YAML frontmatter that helps with discovery, confidence tracking, and surfacing contradictions:
+
+```yaml
+title: "Page Title"
+tags: [tag1, tag2]
+date: YYYY-MM-DD
+source_count: N
+status: active | superseded | archived
+confidence: high | medium | speculative
+contradictions:
+ - "Source A says X, Source B says Y"
+open_questions:
+ - "Unanswered question"
+```
+
+The `confidence` field is especially important — it distinguishes well-sourced claims from early speculation, letting readers (and the LLM itself) calibrate trust.
+
+## Three Operations
+
+An LLM wiki supports three core operations:
+
+### Ingest
+
+Fetch a source into `raw/`, then ripple-update all affected wiki pages. A single ingest may touch a dozen pages — creating new entity pages, updating summaries, adding cross-references, and flagging new contradictions. The index and log are updated automatically.
+
+### Query
+
+Search the wiki and answer questions with citations. Because the wiki is structured markdown with frontmatter, the LLM can cite specific pages and confidence levels rather than hallucinating.
+
+### Lint
+
+Maintenance pass: fix broken links, flag contradictions, find orphan pages, and mark stale content. Think of it as the wiki's CI pipeline.
+
+## Three Key Files
+
+Every LLM wiki has three essential files:
+
+### `index.md` — Content Catalog
+
+A structured listing of every wiki page, organized by category, each with a one-line summary. Updated on every ingest. This is the table of contents for both humans and LLMs.
+
+### `log.md` — Activity Log
+
+Append-only chronological record. Every ingest, query, and lint gets an entry:
+
+```markdown
+## [2026-04-10] ingest | "Attention Is All You Need"
+Added source summary. Updated pages: Transformers, Self-Attention, Google Brain.
+New entity page: Ashish Vaswani.
+```
+
+### `CLAUDE.md` — Schema
+
+The instruction set for the LLM. Defines directory structure, naming conventions, frontmatter fields, cross-reference rules, and domain-specific behavior. Lives at the repo root so any LLM agent can find it.
+
+## Page Content Structure
+
+Each wiki page follows a consistent structure:
+
+- **Summary** — what this page is about, in 2-3 sentences
+- **Key Claims** — falsifiable, citable assertions drawn from sources
+- **Connections** — wikilinks to related pages (symmetric: if A links B, B links A)
+- **Contradictions** — conflicts between sources, pre-flagged so readers see disagreements upfront
+
+This structure makes every page useful both to human readers and to LLMs doing follow-up queries.
+
+## How WikiHub Maps to This Format
+
+WikiHub is the hosting platform for LLM wikis — "GitHub for LLM wikis." Here is how the pieces connect:
+
+- **Git-backed storage** — every wiki is a bare git repo, so you get versioning, diffs, and the full history of how knowledge evolved
+- **Markdown rendering** — WikiHub renders your wiki pages with wikilinks (`[[Page Name]]`), KaTeX math, footnotes, and Obsidian-style embeds
+- **API + MCP access** — LLM agents can read and write pages via the REST API or MCP server, making automated ingest and lint operations straightforward
+- **Frontmatter support** — visibility, tags, and custom fields in frontmatter are preserved and used for access control and discovery
+- **Public + private pages** — raw sources can be kept private while wiki pages are public, matching the three-layer architecture naturally
+
+When you create a wiki on WikiHub, you are creating a git repo ready to hold this structure. Your LLM agent pushes pages via the API or git, and WikiHub handles rendering, search, and discovery.
+
+## Further Reading
+
+- [Karpathy's original gist](https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f) — the canonical description of the format
+- [[Getting Started]] — create your first WikiHub wiki
+- [[Agent Integration]] — connect your LLM agent to WikiHub
+- [[API Reference]] — full endpoint reference for programmatic access
\ No newline at end of file
index.md
new file mode 100644
index 0000000..fde6f4e
@@ -0,0 +1,25 @@
+---
+bio: GitHub for LLM wikis. The archive for humanity.
+title: WikiHub
+visibility: public
+---
+
+## Community
+
+- [[Featured Wikis]] — curated picks from the community
+- **[Startup Ideas](/@jacobcole/jacobcole/ideas/Startup%20Ideas)** by @jacobcole — startup concepts and explorations
+- **[LLM Wiki Ecosystem](/@jacobcole/llm-wiki-ecosystem)** by @jacobcole — research on tools implementing Karpathy's LLM Wiki pattern
+
+## Documentation
+
+- [[Getting Started]] — create an account, your first wiki, writing pages
+- [[LLM Wiki Format]] — the Karpathy LLM Wiki format: directory structure, frontmatter, operations, and how WikiHub hosts it
+- [[Agent Integration]] — MCP server, REST API, content negotiation, git access
+- [[API Reference]] — full endpoint reference
+- [[Examples]] — example wiki ideas and what makes wikihub different
+
+## Philosophy
+
+- [[philosophy/Core Beliefs]] — why wikihub exists, individual voice over consensus, the archive outlives the platform
+- [[philosophy/The Librarian]] — the Snow Crash vision, an AI daemon that has read every public wiki
+- [[philosophy/Onboarding Vision]] — knowledge-first onboarding, the Karpathy moment
\ No newline at end of file
philosophy/Core Beliefs.md
new file mode 100644
index 0000000..08a67ab
@@ -0,0 +1,76 @@
+---
+visibility: public
+---
+
+# Core Beliefs
+
+The living document for wikihub's core beliefs, vision, and why it exists.
+
+---
+
+## Shared Mind for Humanity
+
+Wikipedia proved that strangers could build an encyclopedia together. But it enforced a single voice — "neutral point of view" — and in doing so, flattened the richest parts of human knowledge: perspective, experience, dissent, intuition.
+
+wikihub is building something different: a **shared mind** where every contributor keeps their own voice.
+
+Imagine a doctor's wiki on chronic pain, a patient's wiki on chronic pain, and a researcher's wiki on chronic pain — all interlinked, all forkable, all searchable as one corpus. No edit wars. No lowest-common-denominator consensus. Just the full texture of human understanding, made navigable.
+
+The shared mind isn't one brain. It's a network of brains — each wiki a neuron, each wikilink a synapse, each fork a new branch of thought. The platform's job is to make the whole network traversable while preserving each node's integrity.
+
+### Why AI makes this possible now
+
+The shared mind was always the dream. What makes it achievable now is that AI agents can do the maintenance work that killed every previous attempt:
+
+- **Cross-referencing** — An agent can read 10,000 wikis and surface connections no human would find
+- **Translation** — Not just languages, but between expertise levels, jargon, and frameworks
+- **Maintenance** — The Karpathy LLM Wiki pattern means wikis stay alive instead of rotting. Agents ingest new sources, update cross-references, flag contradictions
+- **Curation** — [[The Librarian]] vision: an AI guide that knows the full constellation and can lead you through it
+
+The previous generation of wikis failed because maintenance was a human bottleneck. When AI handles the gardening, humans can focus on what they're uniquely good at: having experiences, forming opinions, and sharing hard-won knowledge.
+
+### What this looks like at scale
+
+- A student forks a professor's wiki and annotates it with their own understanding
+- An AI agent ingests a research paper and ripple-updates 15 wikis that cite related work
+- A practitioner's wiki on a medical condition ranks alongside the clinical wiki, because lived experience is knowledge too
+- The full graph of human knowledge becomes queryable: "show me every perspective on X" instead of "show me the consensus on X"
+
+This is wikihub's north star: not a single source of truth, but a **constellation of truths** that together form humanity's shared mind.
+
+---
+
+## Humanity's Archive
+
+The archive for humanity isn't one library with one voice. It's a constellation of individual knowledge bases — each one a person's understanding of their corner of the world — that can be traversed, forked, and cross-referenced.
+
+Wikipedia chose consensus. One article per topic, edit wars to resolve disagreements. That works for encyclopedic facts but kills perspective, voice, and the kind of knowledge that comes from lived experience.
+
+wikihub is the opposite. Every person gets their own archive. Karpathy's wiki about transformers is different from your wiki about transformers. Both are valuable. Neither replaces the other. The platform's job is to make the full constellation navigable.
+
+### Why this matters now
+
+The Karpathy moment — when the most respected ML researcher in the world chose to publish his knowledge as flat markdown in a git repo instead of a blog, book, or course — validated the format. Markdown files in a git repo IS the right shape for knowledge in the AI era. Not databases, not CMS, not Notion.
+
+But GitHub is a code host, not a knowledge host. No rendering, no access control, no search, no social layer. There's a gap between "I want a Karpathy-style knowledge base" and "I can publish one." wikihub fills that gap.
+
+### The agent angle is the moat
+
+Anyone can build a wiki host. The `.wikihub/acl` + MCP + content negotiation + `AGENTS.md` onboarding surface means agents are first-class citizens. An agent can sign up, create a wiki, publish pages, read other wikis, and discover knowledge — all without a browser. That's what makes wikihub not just "another wiki platform" but the knowledge layer for the agent era.
+
+## Core beliefs
+
+- **Shared mind, individual voice.** The constellation of wikis IS the shared mind. No single wiki needs to be complete — the network is the knowledge.
+- **Infrastructure, not app.** wikihub is the memory layer + git host. Agents are clients. The site stands alone and serves any LLM/agent.
+- **Individual voice over consensus.** Every person's archive is theirs. No edit wars, no "neutral point of view" policy. Voice is a feature.
+- **Separate repos, seamless integration.** A coding agent should need <50 lines to fully operate a wiki.
+- **YAGNI.** Ship without anti-abuse machinery, without comments, without collaborative editing. Iterate reactively.
+- **API for writes, git pull for reads.** Same split as listhub.
+- **Read liberally, write conservatively.** Postel's Law for frontmatter compatibility.
+- **Trust the agent era on velocity.** No time estimates for coding-agent work.
+- **Attribution as currency.** When [[The Librarian]] cites your wiki, that's the reward. Make knowledge findable and citable.
+- **The archive outlives the platform.** Every wiki is a git repo you can clone and take with you. If wikihub disappears, the knowledge survives.
+
+---
+
+*See also: [[The Librarian]] for the AI vision, [[Onboarding Vision]] for how new users discover wikihub.*
\ No newline at end of file
philosophy/Onboarding Vision.md
new file mode 100644
index 0000000..23c0c0f
@@ -0,0 +1,45 @@
+---
+title: Onboarding Vision
+visibility: public
+---
+
+# Onboarding Vision
+
+Karpathy-informed onboarding — knowledge-first, not account-first.
+
+---
+
+The Karpathy LLM wiki signal: the most respected ML researcher chose flat markdown in a git repo over a blog, book, or course. This validated the format and created demand with no supply. wikihub fills the gap — but the onboarding should feel like "publish your knowledge" not "create an account on a platform."
+
+## Current flow (generic SaaS)
+
+Sign up → empty dashboard → "Create wiki" button → fill form → empty wiki → now what?
+
+## Proposed flow (knowledge-first)
+
+**Arrival:** See beautiful rendered wikis (official wiki + featured content) → "I want mine to look like that"
+
+**Three entry points:**
+
+1. **"Drop your files"** — drag a folder of .md files, get a wiki instantly. Account created behind the scenes if needed.
+2. **"Start from a template"** — pick the Karpathy skeleton (schema.md, index.md, log.md, wiki/, raw/) or a blank wiki.
+3. **"Connect your vault"** — `git remote add wikihub ...` + `git push` for Obsidian users.
+
+**Immediate payoff:** Content is live and rendered within 60 seconds of arriving. Personal wiki exists, `index.md` is your profile. You're not "setting up an account," you're *publishing*.
+
+**Discovery comes after creation:** Explore, star, fork happen once there's content in the system. Not the entry point.
+
+## Concrete implications
+
+- **Landing page:** Show a live preview — "here's what your wiki will look like" with a dogfood wiki rendered as the example. CTA is "publish" not "sign up."
+- **Signup:** Nearly invisible. Drag-and-drop or git push creates the account as a side effect. For web: username + optional email, done, you're in the editor. For API: already one POST.
+- **Post-signup:** Land in personal wiki's `index.md` editor, not an empty dashboard. Template pre-filled with Karpathy-style skeleton. Message is "write something" not "configure your account."
+- **`/explore`:** Where the official wiki earns its keep. New users who aren't ready to publish yet browse real content — curated picks, most-starred, official docs. The "what is this place" moment.
+
+## Why this matters
+
+The audience (Karpathy-gist wave, Obsidian vault owners, ML researchers) already has content. They don't need to be convinced to write — they need a place to put what they've already written. The onboarding should be a funnel from "I have files" to "they're live" with minimal friction. Account setup, ACL configuration, and social features are all things that happen *after* the first publish, not before.
+
+---
+
+*See also: [[Core Beliefs]] for the foundational philosophy, [[The Librarian]] for the AI vision.*
philosophy/The Librarian.md
new file mode 100644
index 0000000..57d5114
@@ -0,0 +1,88 @@
+---
+title: The Librarian
+visibility: public
+---
+
+# The Librarian
+
+Inspired by Neal Stephenson's *Snow Crash* — an AI daemon with instant recall of the entire library, used as a conversational research partner. Not a search engine. A guide.
+
+---
+
+## Literary references
+
+### Snow Crash — The Librarian (primary inspiration)
+
+Neal Stephenson's *Snow Crash* features "The Librarian" — an AI daemon in the Metaverse with instant recall of the entire Library of Congress (merged with the CIA into the "Central Intelligence Corporation"). The protagonist Hiro uses the Librarian as a conversational research partner. Key properties:
+
+- Conversational, not search-based. Hiro doesn't type queries — he talks, follows threads, goes on tangents, and the Librarian adjusts.
+- Has access to everything. The entire corpus.
+- A daemon, not a person. Explicitly software — no pretense of being human. A tool with personality.
+- Connects dots the human can't. Hiro discovers the Snow Crash virus by following threads through ancient Sumerian linguistics, neurolinguistics, and modern drug culture — a path he couldn't have found alone.
+- The CIC is an information marketplace. "Stringers" (contributors) get paid when their information is used. The Library isn't Wikipedia — it's a network of individual contributors whose knowledge has economic value.
+
+### Ready Player One — Halliday's Journals + The Curator
+
+- Halliday's Journals: a public archive on the planet Incipio in the OASIS — one person's entire life and knowledge rendered as explorable rooms. Every film, game, book he ever saw is archived with metadata.
+- The Curator (Ogden Morrow): secretly the co-creator, disguised as a robotic "Jeeves-like" librarian who helps visitors navigate. Has complete knowledge but guides toward discovery rather than giving direct answers.
+- Anorak's Almanac: the downloadable version — Halliday's journal as a portable PDF.
+
+---
+
+## How this maps to wikihub
+
+| Literary concept | wikihub equivalent |
+|---|---|
+| Halliday's Journals / Library of Congress | Each user's wiki — a person's complete knowledge archive |
+| Planet Incipio / The CIC | The wikihub platform — where all archives live and are discoverable |
+| The Librarian / The Curator | An AI agent that navigates across all public wikis, guides discovery, finds cross-wiki connections |
+| Anorak's Almanac | `git clone` / ZIP download — the portable, offline version |
+| The Easter Egg Hunt | Cross-wiki connections — wikilinks, shared tags, knowledge graphs |
+| Stringers getting paid | Attribution as currency — when the Librarian references your wiki, you get visibility, stars, forks |
+
+---
+
+## The "Humanity's Archive" vision
+
+Wikipedia chose consensus — one voice, one article, edit wars. That works for facts but kills perspective.
+
+wikihub is the opposite: every person gets their own archive. Karpathy's wiki isn't "the article about transformers" — it's Karpathy's understanding. Someone else's wiki about the same topic would be different. Both valuable. Neither replaces the other.
+
+The archive for humanity isn't one library — it's a constellation of individual knowledge bases that can be traversed, forked, and cross-referenced. The platform's job is to make that constellation navigable.
+
+---
+
+## Naming / branding options
+
+| Layer | Option A | Option B | Option C |
+|---|---|---|---|
+| Platform | wikihub | The Library | Humanity's Archive |
+| AI agent | The Librarian | The Curator | The Archivist |
+| Individual exports | Almanac | Journal | Archive |
+| Curated showcase | The Journals | Featured | The Collection |
+| Vision tagline | "The archive for humanity" | "Your knowledge, humanity's library" | "Every mind, an archive" |
+
+**Strong contenders for the platform name:**
+- **wikihub** — functional, already in use, domain locked (wikihub.md)
+- **The Library** — evokes Snow Crash directly, simple, but maybe too generic
+- **Humanity's Archive** — the grand vision name, better as a tagline than a product name
+
+**Strongest combo:** Keep wikihub as the product name. Use "Humanity's Archive" as the vision/tagline. Name the AI agent "The Librarian."
+
+---
+
+## The Librarian as a product feature
+
+The Librarian would be wikihub's AI agent — accessible via `@librarian` on the platform or in conversational mode via Cmd+K:
+
+- Has read access to all public wikis on the platform
+- Conversational — ask questions, it pulls from across wikis and synthesizes
+- Finds cross-wiki connections individual authors can't see
+- Credits sources — "according to @karpathy's transformer wiki..." (attribution built in)
+- The one entity that has read everything on the platform
+- Explicitly a tool/daemon, not pretending to be human — but with personality
+- Powered by the MCP surface — uses the same API any agent uses, just the best-informed one
+
+---
+
+*See also: [[Core Beliefs]] for the foundational philosophy, [[Onboarding Vision]] for the user journey.*