2,625 tagged knowledge-base, measured the same way as everything else here.
Browse within: local-first 37markdown 36obsidian 34karpathy 27knowledge-graph 27embeddings 24hybrid-search 24fastapi 17llm-wiki 17agent-memory 16knowledge-management 16best-practices 11documentation 11wiki 11
HKUST-KnowComp/DeepRefine-Skill
Skill Claude CodeCodex
Agent-native DeepRefine refinement loop — same control flow as DeepRefine.refine(), graphify search instead of FAISS, session LLM, dry-run review before approved graph writes.
HKUST-KnowComp/DeepRefine-Skill
Skill Claude CodeCodex
Claude Code adapter for the DeepRefine agent-native refinement loop. Use when the user invokes /deeprefine, or asks to refine, diagnose, review, or apply changes to a Graphify / LLM-Wiki knowledge graph. Must follow the canonical DeepRefine skill rules and stop for review before graph writes.
Instructions file
Claude Code instructions for frankchu91/mindbase-llm-wiki, covering mindbase — north star for claude code, product thesis (one paragraph), the 3-layer architecture (from karpathy), the 3 operations (from karpathy) and special files (from karpathy).
Plugin Claude Code
Karpathy's LLM-Wiki pattern as a Claude Code plugin. Build and maintain a persistent, AI-tended research wiki from sources, capture daily thoughts, query what you know, lint for orphans and contradictions. Operations: /mb:init, /mb:contribute, /mb:build, /mb:ask, /mb:lint, /mb:status, /mb:research, /mb:daily-brief…
MCP server Claude CodeCodexCursor +2
MCP server for MindBase — query and write to your wiki from Claude Desktop, Cursor, Windsurf, Cline, or any MCP-compatible AI client. Runs locally from the mindbase-mcp npm package.
Agent
Sub-agent dispatched by /mb:build to synthesize context.md from sources/. Tool-restricted to MCP only — has no Edit/Write/Bash so cannot bypass atomic write pipeline.
Agent
Sub-agent dispatched by /mb:lint. Read-only — has Read + lint MCP tools but no Edit/Write/Bash. Surfaces orphans, contradictions, stale claims, gaps. Suggests fixes; doesn't apply them.
Agent
One-time sub-agent that runs /mb:migrate. Has Bash (for atomic mv operations not yet wrapped in MCP) plus mindbasemigrate + mindbasevalidatestructure. Should be invoked exactly once per legacy project.
Command
Rebuild context.md for a project from sources/. Usage: /mb:build [-p ] [project-id].
Command
Add a thought, source, or paste. Use --project/-p to target another project without switching current. Usage: /mb:contribute [-p ] [--mode daily|concept|daily+concept].
Command
Scaffold a new MindBase project (v2 layout). Usage: /mb:init [template] [-- mission ...].
Hook
Runs when a session starts on startup, clear and compact, running an inline shell check. From frankchu91/mindbase-llm-wiki.
Skill Claude CodeCodex
Maintain a persistent AI-tended research wiki. Activates when the user wants to ingest a source (PDF, URL, article, paste), capture a thought, build the wiki context, query what's already known, lint for orphans/contradictions, or maintain page conventions. Phrases include "add this to my wiki", "today I decided…"…
Hook Claude Code
Runs when a session starts on startup and resume, executing cloud-session-setup.sh. From zby/commonplace.
Settings file Claude Code
Agent settings declaring 1 hook event (SessionStart) and 1 allowed tools.
Instructions file CodexOpenCode
AGENTS.md instructions for zby/commonplace, covering commonplace, repository overview, kb goals and scope, purpose and scope.
Instructions file
Claude Code instructions for zby/commonplace, a project described as: The theory of LLM wikis, running as one. A framework for agent-operated knowledge: typed, linked, review-gated markdown your agents execute.
Skill Claude CodeCodex
Use when asked to analyse, review, or refresh an external agent runtime, orchestration system, agent operating layer, agent memory/knowledge/context-engineering system, or narrower model-dependent operational mechanism from inspectable sources.
Skill Claude CodeCodex
Experimentally revise a note with fresh critic, author, synthesizer, and blind-judge agents while retaining the incumbent as fallback.
Skill Claude CodeCodex
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
Agent
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
Agent
Compare two outputs WITHOUT knowing which skill produced them.
Agent
Evaluate expectations against an execution transcript and outputs.
Instructions file CodexOpenCode
A set of project instructions for coding agents working on Klear-Team-Brain. It points agents to documentation about the architecture, data flow, deployment, and release changes.
At most 3 mods per repository are shown here — the rest are on their repository pages: