docs(brand): flip display prose to "LLM Wiki Bridge" across 14 files

2233admin · 2233admin · commit 8bb3be07cd7e · 2026-04-21T01:19:19.000+08:00
Technical slug (@obsidian-llm-wiki/mcp, github.com/2233admin/obsidian-llm-wiki,
serverInfo.name, clone paths, kb_meta path) unchanged. Prose flipped in
README H1+pitch, CLAUDE.md H1, setup/setup.ps1 top+echo banners,
REQUIREMENTS H1 + install prefix, .outreach/{show-hn,x-thread},
viewer title+h1, demo-vault + creatures READMEs, WHY_NOT_JUST_GREP,
gif-script, dogfood-log.sh header, vault-mind.example.yaml header.

Closes step 5/6 of rebrand workflow (see feedback_rebrand_workflow memory).
Negative grep verified: 30 residual slug mentions, all technical
(URLs, npm scope, clone paths, serverInfo.name, script ROOT).
Karpathy concept attribution preserved (23 mentions untouched).
diff --git a/.outreach/show-hn-draft.md b/.outreach/show-hn-draft.md
@@ -1,6 +1,6 @@
 # Draft: Show HN post
 
-**Title**: `Show HN: Karpathy's LLM Wiki, without the Postgres (obsidian-llm-wiki v1.0)`
+**Title**: `Show HN: Karpathy's LLM Wiki, without the Postgres (LLM Wiki Bridge v1.0)`
 
 **URL**: https://github.com/2233admin/obsidian-llm-wiki
 
@@ -10,7 +10,7 @@
 
 When Karpathy published his LLM Wiki gist in April, the comment thread immediately filled with variations that added things the gist explicitly said weren't needed at small scale — mostly embeddings and vector DBs.
 
-obsidian-llm-wiki is a deliberately orthodox implementation:
+LLM Wiki Bridge is a deliberately orthodox implementation:
 
 - **Three operations**, as the gist describes: Ingest / Query / Lint. Implemented as MCP tools so any MCP client (Claude Code, Cline, etc.) can drive them.
 - **No embeddings at moderate scale.** Filesystem adapter uses ripgrep. Embeddings/memU/pgvector are *optional* adapters — you can ignore them entirely and still have a working system.
diff --git a/.outreach/x-thread.md b/.outreach/x-thread.md
@@ -28,7 +28,7 @@ Why not gbrain? Different use case.
 
 gbrain = Postgres + pgvector + dream cycles. Built for 10k+ files and long-running agents.
 
-obsidian-llm-wiki = plain markdown + Obsidian UI + MCP. Built for your personal vault, your laptop, your Claude Code session.
+LLM Wiki Bridge = plain markdown + Obsidian UI + MCP. Built for your personal vault, your laptop, your Claude Code session.
 
 **Tweet 4 (collectors)**
 Ships with 10 source collectors that dump directly to your vault:
diff --git a/.planning/REQUIREMENTS.md b/.planning/REQUIREMENTS.md
@@ -1,4 +1,4 @@
-# Requirements: obsidian-llm-wiki v2 (shipability)
+# Requirements: LLM Wiki Bridge v2 (shipability)
 
 **Defined:** 2026-04-20 (v2 — ship-ready pivot)
 **Core Value:** Your markdown vault becomes a 6-persona virtual team for Claude Code / Codex / OpenCode / any agent. MCP + compiler are plumbing; personas are the product.
@@ -31,7 +31,7 @@ Each persona: 1 markdown file in `skills/`, < 150 lines, calls MCP tools via Cla
 ### Install UX (Layer 1 — distribution)
 
 - [ ] **I-01**: Paste-to-install prompt for Claude Code:
-  `Install obsidian-llm-wiki: git clone --depth 1 https://github.com/2233admin/obsidian-llm-wiki ~/.claude/skills/vault-wiki && cd ~/.claude/skills/vault-wiki && ./setup` (mirror gstack).
+  `Install LLM Wiki Bridge: git clone --depth 1 https://github.com/2233admin/obsidian-llm-wiki ~/.claude/skills/vault-wiki && cd ~/.claude/skills/vault-wiki && ./setup` (mirror gstack).
 - [ ] **I-02**: `./setup` script: detects vault path (env VAULT_PATH or CWD check), writes `.mcp.json` snippet, appends to CLAUDE.md the 6 persona list.
 - [ ] **I-03**: Multi-agent support: `./setup --host claude|codex|opencode` installs skill in the right dir (match gstack auto-detect table).
 - [ ] **I-04**: Team mode: `./setup --team` puts skill as a dev dep + auto-update check. Deferred to v2.1 if time is tight.
diff --git a/CLAUDE.md b/CLAUDE.md
@@ -1,4 +1,4 @@
-# CLAUDE.md — obsidian-llm-wiki
+# CLAUDE.md — LLM Wiki Bridge
 
 ## What This Is
 
diff --git a/README.md b/README.md
@@ -1,12 +1,12 @@
-# obsidian-llm-wiki
+# LLM Wiki Bridge
 
 Your markdown vault, compiled into a 6-persona team for any agent.
 
 ![demo](docs/gif/demo.gif)
 
 Your vault has 500 notes. You forget half of them. Your AI agent cannot read them. Every morning you spend 20 minutes re-finding what you already knew.
 
-obsidian-llm-wiki solves this. It compiles your markdown vault -- wikilinks, aliases, tags, frontmatter -- into an MCP server your agent calls directly. The agent does not guess. It searches, reads, and cites with full vault context.
+LLM Wiki Bridge solves this. It compiles your markdown vault -- wikilinks, aliases, tags, frontmatter -- into an MCP server your agent calls directly. The agent does not guess. It searches, reads, and cites with full vault context.
 
 ---
 
diff --git a/creatures/README.md b/creatures/README.md
@@ -1,7 +1,7 @@
 # spec-A Output: 6 Vault Personas
 
 Output of `librarian-author` on MiniMax-M2.7-highspeed via KT.
-Built for `obsidian-llm-wiki` v2.
+Built for `LLM Wiki Bridge` v2.
 
 ## Personas
 
diff --git a/docs/WHY_NOT_JUST_GREP.md b/docs/WHY_NOT_JUST_GREP.md
@@ -14,7 +14,7 @@ Example: your vault has a note titled `A3: KV-Cache Optimization` with alias `kv
 
 ## 2. Grep is stateless. This is compiled.
 
-Grep scans on every invocation. The first query costs the same as the 1,000th. obsidian-llm-wiki compiles your vault once. On a 554-note vault, it produces a concept graph with 2,507 edges in under 3 seconds. Every subsequent query reuses that graph. The agent calls `vault.backlinks` and gets the full backlink map without rescanning the filesystem.
+Grep scans on every invocation. The first query costs the same as the 1,000th. LLM Wiki Bridge compiles your vault once. On a 554-note vault, it produces a concept graph with 2,507 edges in under 3 seconds. Every subsequent query reuses that graph. The agent calls `vault.backlinks` and gets the full backlink map without rescanning the filesystem.
 
 ---
 
diff --git a/docs/gif-script.md b/docs/gif-script.md
@@ -7,7 +7,7 @@ Show Claude Code window. Cursor is on an empty line in a project directory.
 Type (animate character by character):
 
 ```
-Install obsidian-llm-wiki: git clone --depth 1 https://github.com/2233admin/obsidian-llm-wiki ~/.claude/skills/vault-wiki && cd ~/.claude/skills/vault-wiki && ./setup
+Install LLM Wiki Bridge: git clone --depth 1 https://github.com/2233admin/obsidian-llm-wiki ~/.claude/skills/vault-wiki && cd ~/.claude/skills/vault-wiki && ./setup
 ```
 
 Press Enter. Show brief terminal output scrolling (setup runs, completes).
diff --git a/examples/demo-vault/README.md b/examples/demo-vault/README.md
@@ -1,6 +1,6 @@
 # Demo Vault
 
-10 synthetic markdown notes about AI engineering topics, used to demo the obsidian-llm-wiki personas. Points the `/vault-*` skills at this vault for a working demonstration.
+10 synthetic markdown notes about AI engineering topics, used to demo the LLM Wiki Bridge personas. Points the `/vault-*` skills at this vault for a working demonstration.
 
 ## How to point MCP server at this vault
 
diff --git a/scripts/dogfood-log.sh b/scripts/dogfood-log.sh
@@ -1,7 +1,7 @@
 #!/usr/bin/env bash
 # dogfood-log.sh — append a single usage record to .dogfood.log
 #
-# Purpose: Track real daily usage of obsidian-llm-wiki so we can see at Week 12
+# Purpose: Track real daily usage of LLM Wiki Bridge so we can see at Week 12
 # which features are used 100+ times (keep) vs 0 times (cut).
 #
 # Usage:
diff --git a/setup b/setup
@@ -1,5 +1,5 @@
 #!/usr/bin/env bash
-# obsidian-llm-wiki setup -- install skill + configure MCP
+# LLM Wiki Bridge setup -- install skill + configure MCP
 #
 # Copies a curated allowlist into the host's skills directory.
 # ~1.7 MB install (vs 64 MB of repo) -- ships a pre-bundled MCP server
@@ -101,7 +101,7 @@ run cp "$SCRIPT_DIR/mcp-server/package.json" "$SKILLS_DIR/mcp-server/package.jso
 INSTALL_PATH="$SKILLS_DIR/mcp-server/bundle.js"
 
 echo ""
-echo "obsidian-llm-wiki installed to:"
+echo "LLM Wiki Bridge installed to:"
 echo "  $SKILLS_DIR"
 echo ""
 echo "Next steps:"
diff --git a/setup.ps1 b/setup.ps1
@@ -1,4 +1,4 @@
-# obsidian-llm-wiki setup (PowerShell)
+# LLM Wiki Bridge setup (PowerShell)
 # Usage: .\setup.ps1 [-VaultHost <name>] [-DryRun]
 #
 # Copies a curated allowlist into the host's skills directory.
diff --git a/vault-mind.example.yaml b/vault-mind.example.yaml
@@ -1,4 +1,4 @@
-# obsidian-llm-wiki -- runtime configuration example
+# LLM Wiki Bridge -- runtime configuration example
 #
 # Copy this file to `vault-mind.yaml` (git-ignored) and edit paths/flags
 # for your machine. Or skip the yaml entirely and set
diff --git a/viewer/index.html b/viewer/index.html
@@ -3,13 +3,13 @@
 <head>
   <meta charset="UTF-8" />
   <meta name="viewport" content="width=device-width, initial-scale=1.0" />
-  <title>obsidian-llm-wiki — Concept Graph Viewer</title>
+  <title>LLM Wiki Bridge — Concept Graph Viewer</title>
   <link rel="stylesheet" href="viewer.css" />
 </head>
 <body>
 
 <header>
-  <h1>obsidian-llm-wiki — Concept Graph</h1>
+  <h1>LLM Wiki Bridge — Concept Graph</h1>
   <div class="paste-area">
     <label for="graph-json">Paste your graph.json:</label>
     <textarea id="graph-json" placeholder="Paste the output of concept_graph.py here..."></textarea>

Original file line number	Diff line number	Diff line change
`@@ -1,4 +1,4 @@`
`1`		`-# CLAUDE.md — obsidian-llm-wiki`
	`1`	`+# CLAUDE.md — LLM Wiki Bridge`
`2`	`2`
`3`	`3`	`## What This Is`
`4`	`4`
Original file line number	Diff line number	Diff line change
`@@ -1,7 +1,7 @@`
`1`	`1`	`#!/usr/bin/env bash`
`2`	`2`	`# dogfood-log.sh — append a single usage record to .dogfood.log`
`3`	`3`	`#`
`4`		`-# Purpose: Track real daily usage of obsidian-llm-wiki so we can see at Week 12`
	`4`	`+# Purpose: Track real daily usage of LLM Wiki Bridge so we can see at Week 12`
`5`	`5`	`# which features are used 100+ times (keep) vs 0 times (cut).`
`6`	`6`	`#`
`7`	`7`	`# Usage:`