new file mode 100644
index 0000000..e5a2282
@@ -0,0 +1,152 @@
+---
+title: Related Projects, People, and Landscape
+visibility: public
+---
+
+# Related Projects, People, and Landscape
+
+The LLM wiki / personal knowledge base space as of April 2026. People building in the space, their projects, key tweets, and how they relate to WikiHub.
+
+---
+
+## Key People
+
+### Andrej Karpathy
+- **Role:** Coined the "LLM Wiki" pattern (April 2, 2026)
+- **Twitter:** [@karpathy](https://x.com/karpathy)
+- **Key artifact:** [LLM Wiki gist](https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f) — 3-layer architecture (raw sources / wiki / schema)
+- **Key tweet:** ["LLM Knowledge Bases"](https://x.com/karpathy/status/2040016856015839590) — 12M+ views, started the wave
+- **Quote:** "Obsidian is the IDE, the LLM is the programmer, the wiki is the codebase"
+- **Impact:** ~8,200 stars on the gist in 12 hours. Spawned 20+ open-source implementations within a week.
+- **Relevance to WikiHub:** Karpathy's pattern is the foundation. WikiHub is the hosted publishing layer that the pattern is missing — a place to publish, share, and clone LLM wikis.
+
+### Farza Majeed (@FarzaTV)
+- **Role:** Built Farzapedia — the first high-profile personal LLM wiki
+- **Twitter:** [@FarzaTV](https://x.com/FarzaTV)
+- **Background:** Founder of buildspace (a16z-backed). Currently building Tidbit with Sharif Shameem.
+- **Key tweet:** ["This is Farzapedia"](https://x.com/FarzaTV/status/2040563939797504467) — April 4, 2026. 1.7M views, 4,633 likes, 5,826 bookmarks.
+- **What he built:** 2,500 entries from diary + Apple Notes + iMessage → 400+ interlinked wiki articles covering friends, startups, philosophy, anime, relationships
+- **Data sources:** Day One journal, Apple Notes (exported HTML/TXT/MD), iMessage (extracted from `~/Library/Messages/chat.db`)
+- **Published skill:** [personal_wiki_skill.md](https://gist.github.com/farzaa/c35ac0cfbeb957788650e36aabea836d) — 505-line Claude Code skill with commands: `/wiki ingest`, `/wiki absorb`, `/wiki query`, `/wiki cleanup`, `/wiki breakdown`, `/wiki status`
+- **Key insight:** Auto-generated a "Relationship History" article from diary + iMessage — sensitive enough that Farzapedia will never be public. This is the core privacy problem WikiHub's per-page visibility solves.
+- **Landing page:** [farza.com/knowledge](https://farza.com/knowledge) — interest-gauge page, not yet a product
+- **Relevance to WikiHub:** Farzapedia proves the demand. The fact that it can't be made public (one sensitive article poisons the whole vault) is exactly the problem WikiHub's `<!-- private -->` blocks and per-file visibility solve.
+
+### Naman Ambavi (@namanambavi)
+- **Role:** Built Namanopedia / mylife.wiki — an LLM wiki generator for anyone
+- **Twitter:** [@namanambavi](https://x.com/namanambavi)
+- **Background:** CEO of [Oximy](https://oximy.com) (YC W26) — "Know how AI is actually used across your company." Previously COO at Induced AI, founder of Easo Care.
+- **Key tweet:** ["This is Namanopedia"](https://x.com/namanambavi/status/2041999388680319226) — April 8, 2026. 8,871 views.
+- **What he built:** [mylife.wiki](https://mylife.wiki) — paste a name, get a 40-50+ article Wikipedia about that person. AI agent researches the web and compiles articles.
+- **Demo:** [mylife.wiki/naman-ambavi](https://mylife.wiki/naman-ambavi)
+- **How it differs:** Farzapedia/Karpathy are from private data (diary, messages). Namanopedia is from public web data about a person. Different input, same output shape.
+- **Relevance to WikiHub:** Shows that the wiki-as-output format works for public-data-about-a-person too, not just private-data-by-a-person. WikiHub could host both patterns.
+
+### AYi (@AYi_AInotes)
+- **Twitter:** [@AYi_AInotes](https://x.com/AYi_AInotes)
+- **Key tweet:** [Detailed Chinese-language breakdown of Farzapedia](https://x.com/AYi_AInotes/status/2040635298179428371) — translated and analyzed Farza's approach for the Chinese AI community
+- **Relevance:** Shows international interest in the pattern. WikiHub should consider i18n early.
+
+---
+
+## Open-Source Implementations
+
+| Project | Author | Stars | Description |
+|---------|--------|-------|-------------|
+| [karpathy/llm-wiki gist](https://gist.github.com/karpathy/442a6bf555914893e9891c11519de94f) | Karpathy | ~8.2K | The original idea file |
+| [farzaa/personal_wiki_skill.md](https://gist.github.com/farzaa/c35ac0cfbeb957788650e36aabea836d) | Farza | — | 505-line Claude Code skill |
+| [kothari-nikunj/llm-wiki](https://github.com/kothari-nikunj/llm-wiki) | Nikunj Kothari | — | Most complete OSS version. iMessage extraction, WhatsApp, Apple Notes, Next.js viewer |
+| [SamurAIGPT/llm-wiki-agent](https://github.com/SamurAIGPT/llm-wiki-agent) | SamurAI | — | Drop-in agent for Claude/Codex/Gemini. NetworkX graph viz |
+| [Astro-Han/karpathy-llm-wiki](https://github.com/Astro-Han/karpathy-llm-wiki) | Astro-Han | — | One Agent Skill for Claude/Cursor/Codex |
+| [tashisleepy/knowledge-engine](https://github.com/tashisleepy/knowledge-engine) | tashisleepy | — | Karpathy + Memvid, dual human/machine layer |
+| [kfchou/wiki-skills](https://github.com/kfchou/wiki-skills) | kfchou | — | Claude Code skills (init, ingest, query, lint, update) |
+| [ussumant/llm-wiki-compiler](https://github.com/ussumant/llm-wiki-compiler) | ussumant | — | Claude Code plugin, ~90% context-cost reduction |
+| [swarajbachu/cachezero](https://github.com/swarajbachu/cachezero) | swarajbachu | — | NPM CLI implementing the pattern (Show HN) |
+| [llmwiki.app](https://llmwiki.app/) | — | — | Hosted free open-source implementation via Claude |
+
+---
+
+## Named Personal Wikis (People Who Actually Built One)
+
+| Wiki | Person | Size | Public? | Notes |
+|------|--------|------|---------|-------|
+| **Farzapedia** | Farza Majeed | ~400 articles from 2,500 entries | No ("Relationship History" problem) | Most famous example |
+| **Commonplace** | zby (HN: zbyforgotpass) | ~30+ system reviews | [Yes](https://zby.github.io/commonplace/) | Wiki about LLM agent memory systems, maintained by agents (recursive) |
+| **Grimoire** | sprites.app dev | Active commits | [Yes](https://grimoire-pt5.sprites.app/) | AI research, infra, security themes |
+| **Karpathy's own** | Andrej Karpathy | ~100 articles, ~400K words | No | ML research topic, described in gist |
+| **Louis Wang's KB** | Louis Wang | ~100-200 articles | [Writeup only](https://louiswang524.github.io/blog/llm-knowledge-base/) | Best scaling analysis ("index-first ceiling") |
+| **Namanopedia** | Naman Ambavi | 40-50+ articles per person | [Yes](https://mylife.wiki/naman-ambavi) | Public-web-sourced, not personal data |
+| **Jacobpedia** | Jacob Cole | 2,300+ articles from 10,900 entries | No (private, local only) | iMessage + Apple Notes + daily notes + documents |
+
+---
+
+## AI-Generated Public Encyclopedias (Sense 2)
+
+| Project | Builder | Scale | URL |
+|---------|---------|-------|-----|
+| **Grokipedia** | xAI | 885K articles | [grokipedia.com](https://grokipedia.com/) |
+| **Endless Wiki** | vgel | 62K+ pages | [theinfinite.wiki](https://theinfinite.wiki/) |
+| **DeepWiki** | Cognition | 50K+ repo wikis | [deepwiki.com](https://deepwiki.com/) |
+| **WikiGen.ai** | — | — | [wikigen.ai](https://www.wikigen.ai/) |
+| **WikiCrow** | FutureHouse (Stanford) | Protein-coding genes | [wikicrow.ai](https://wikicrow.ai/) |
+
+---
+
+## Wikis FOR LLMs (Sense 3)
+
+| Standard | Adoption | URL |
+|----------|----------|-----|
+| **llms.txt** | 844K+ sites (Anthropic, Cursor, Cloudflare, Vercel) | [llmstxt.org](https://llmstxt.org/) |
+| **DeepWiki** | 50K+ repos | [deepwiki.com](https://deepwiki.com/) |
+| **Semiont** (AI Alliance) | Graph-based, W3C Web Annotation, MCP-exposed | [GitHub](https://github.com/The-AI-Alliance/semiont) |
+| **AgentWiki** | DokuWiki + JSON-RPC | [agentwiki.org](https://agentwiki.org/) |
+
+---
+
+## Human + AI Collaborative Wikis (Sense 5)
+
+- **Semiont** (AI Alliance) — Graph-based knowledge with W3C Web Annotation and MCP
+- **AgentWiki** — Traditional DokuWiki with agent read/write via JSON-RPC
+- **WikiHub** (this project) — Git-backed markdown wikis with per-page visibility, agent API, and clone support
+
+---
+
+## Key Directories / Meta-Resources
+
+| Directory | Focus | Size |
+|-----------|-------|------|
+| [llmstxthub.com](https://llmstxthub.com/) | llms.txt sites (curated, PR-based) | 500+ |
+| [llmtxt.app](https://llmtxt.app/) | llms.txt sites (crawl-based) | ~13,000 |
+| [deepwiki.directory](https://deepwiki.directory/) | Repo wikis | — |
+| [mcp.so](https://mcp.so) | MCP servers | — |
+
+---
+
+## Five-Sense Taxonomy
+
+From our [LLM wiki research](/@jacobcole/architecture/beads-vs-wikihub-architecture):
+
+1. **Personal LLM wiki (Karpathy sense)** — LLM curates a markdown KB for one user
+2. **AI-generated public encyclopedia** — LLM writes articles at scale for public
+3. **Wiki FOR LLMs to consume** — docs shaped for LLM ingestion (llms.txt)
+4. **Wiki ABOUT LLMs** — traditional human-written reference content
+5. **Human + AI collaborative wiki** — shared surfaces where both read/write
+
+WikiHub sits at the intersection of senses 1, 3, and 5.
+
+---
+
+## WikiHub's Position
+
+WikiHub is the missing **publishing and hosting layer** for the Karpathy/Farzapedia pattern:
+
+- Karpathy's pattern gives you the local wiki. WikiHub gives you a URL to share it.
+- Farzapedia proves the privacy problem. WikiHub's per-file visibility and `<!-- private -->` blocks solve it.
+- Namanopedia proves the public-data variant works. WikiHub can host both.
+- The llms.txt ecosystem proves agent-readable web content has demand. WikiHub serves markdown natively via `Accept: text/markdown`.
+
+No existing tool combines: multi-user + browser-editable + per-page visibility + REST API + git clone + agent-native.
+
+---
+
+*Last updated: 2026-04-10*
\ No newline at end of file