Conceptual Primers and Foundations
High-level conceptual breakdowns, foundational ideas, and introductory guides to the LLM Wiki concept.
16 Million Views for a Folder Structure (HR Srinivas)hrsrini.github.ioA conceptual primer analyzing Andrej Karpathy's LLM Wiki pattern and 'idea file' paradigm. Details the three-layer architecture, the Ingest-Query-Lint execution cycle with back-filing query answers, and the historical connection to Vannevar Bush's 1945 Memex—showing how zero-cost AI maintenance solves the 81-year-old associative trail bottleneck.

A conceptual primer based on real-world multi-agent development. Breaks down six core advantages of grounding LLMs in an interconnected Markdown wiki: cross-session persistent memory, linking decisions directly to work tickets, team knowledge democratization, storing agent skills as Markdown, automated maintenance, and local file retrieval speed.
- Andrej Karpathy Wiki: LLM Wiki Concept Guideandrej-karpathy.com
A conceptual breakdown embedded inside the live Karpathy encyclopedia detailing the explicit operating models, file structures, and automated compilation loops that define a production-grade LLM Wiki setup.

A conceptual primer explaining the LLM Wiki pattern. Contrasts standard vector RAG (where the model remains a "tourist") with Karpathy's compiled knowledge approach (where the model acts as a "compiler"), detailing the workflow steps of collection, compilation, maintenance, and querying.
Andrej Karpathy’s LLM Wiki: Full Breakdown and How to Build Your Own (Hari Krishna on Tech in General Substack)nandigamharikrishna.substack.comA comprehensive conceptual primer detailing the limitations of stateless RAG, explaining the three-layer architecture (immutable raw sources, LLM-generated markdown notes, and system schema configuration), outlining the three core operations (Ingest, Query, Lint), and connecting the pattern to Vannevar Bush's Memex.
Beyond LLM Wiki: Build and Query Knowledge Graphs Progressively (URIBurner)linkeddata.uriburner.comAn architectural guide and interactive infographic formalizing the progression from flat-file LLM Wikis to linked data knowledge graphs. Details how agent skills ingest source URLs into schema.org ontologies, compile RDF Turtle graphs into Virtuoso-backed SPARQL endpoints, and query structured entities using semantic inference pragmas.
Building an 'Agent Only' Obsidian Vault (The Thinkers Club)thethinkers.clubA technical architectural guide defining a human-agent sandbox paradigm. Details a methodology for isolating autonomous AI writers inside a dedicated secondary vault to compile a self-organizing wiki of consumed data, preventing context contamination and layout drift in human-authored notes.

An architectural guide exploring how to extend Karpathy LLM Wikis into Semantic Web knowledge graphs. Details how an AI agent skill ingests any document URL into schema.org ontologies and compiles queryable RDF graphs into Virtuoso-backed SPARQL endpoints.

A technical primer exploring how OpenWiki manages memory staleness and drift for evolving codebases. Details the use of a "claims runtime" that connects wiki assertions to versioned code evidence, allowing agents to detect, track, and correct stale knowledge without rebuilding vaults from scratch.

An architectural essay examining the retention of disproven hypotheses and superseded premises in agent memory vaults. Explains why deleting invalidated facts causes agents to re-derive failed approaches, detailing a non-destructive burial pattern with bi-temporal validity bounds, bidirectional supersession links, and consolidation loops that keep the active context core compact.
Commonplacezby.github.ioA technical framework establishing the theory of deploy-time learning for bounded AI observers, detailing semantic distillation operations, and providing automated CLI workspace management skills.
Compounding Knowledge with LLMs: Karpathy's Wiki Pattern in Actionpub.towardsai.netAn architectural analysis connecting Andrej Karpathy's LLM Wiki pattern to Vannevar Bush's historical vision of the Memex. Discusses how LLMs resolve the bottleneck of active knowledge maintenance, detailing layout models, indexing processes, and scalable retrieval strategies for multi-agent environments.
Create Your AI Brain Todaysaranfn.substack.comThe final part of a structural context trilogy pinpointing why knowledge management setups fail at the content layer. Focuses on content architecture optimization, detailing how to engineer high-signal markdown nodes, construct type-safe frontmatter blocks for explicit entity tracking, and implement strict data compression filters to protect agent context windows from token bloat.
Delete Everything, Keep the Graph: LLM-Wiki and OKF (Paul Hackenberger on Medium)paul-hackenberger.medium.comA conceptual primer framing graph engineering as the durable foundation for AI agent memory. Synthesizes Andrej Karpathy's LLM Wiki pattern with Google's Open Knowledge Format (OKF), arguing that while models, prompts, and frameworks are ephemeral, an agent-maintained Markdown graph (pages as nodes, wikilinks as edges, YAML frontmatter as typed properties) is the only persistent asset that compounds over time.

An architectural essay examining knowledge system consistency across five dimensions (logical, semantic, temporal, dependency, provenance) split into state and support layers. Connects Karpathy's LLM Wiki compilation loops with classical Truth Maintenance Systems (TMS) and database provenance semirings to solve silent cascading failures when source premises drift or are retracted.
Give Your AI Unlimited Updated Contextsaranfn.substack.comAn operational architecture guide exploring decoupled, portable text vaults. Outlines the strategic benefits of maintaining plain text markdown knowledge graphs over proprietary, model-siloed memories, detailing how to utilize root configuration files (
CLAUDE.md) and automated background agent loops to keep cross-functional context updated natively.
Inside the LLM Wiki Movement: Build Knowledge That Won't Betray You (Allen on AFFiNE Blog)affine.proA deep conceptual primer examining the LLM Wiki movement, contrasting the four core architectures (single-file, multi-file vault, code-based, and RAG-augmented), discussing hallucination mitigation via confidence tagging, and reviewing platform trade-offs.

The official announcement of OpenWiki Brains, detailing how agents use flat-file local wikis as proactive long-term memory.

A comprehensive architectural guide breaking down Andrej Karpathy's LLM Wiki pattern. Analyzes the three-layer layout (raw sources, agent-maintained markdown graph, schema), details the Ingest-Query-Lint execution cycle and index navigation files, and contrasts compiled persistent memory against stateless runtime RAG.

A conceptual framework published by the Agentic AI Foundation mapping the LLM Wiki pattern to cognitive agent memory architectures. Explains how directory boundaries transform flat markdown files into distinct functional memory layers—mapping schemas to procedural execution rules, logs to episodic records, and cross-linked entity directories to semantic graph memory.

An architectural breakdown of the LLM Wiki pattern contrasting it with standard RAG pipelines, defining the three-layer structure and three operations.
Karpathy's LLM Wiki: The Complete Guide (Agentpedia Codes)antigravity.codesAn exhaustive breakdown analyzing the three-layer architecture, comparing static compilation vs. traditional RAG, and detailing prompt configurations.

A conceptual primer and multi-agent retrospective detailing Karpathy's three-layer architecture, comparing 12 community GitHub implementations, and sharing operational lessons from running a 115-page hosted wiki shared across four autonomous agents over MCP.
LLM Markdown Wiki: A Personal Second Brain Without the Maintenance (Kiryl Bahdanovich on WAVEPILLARS)wavepillars.comA conceptual guide detailing how an LLM-maintained markdown wiki cuts knowledge maintenance overhead. Explores the raw/wiki/schema architecture, Obsidian graph-based hierarchy reviews, Cursor agent sync workflows, and utilizing frontmatter-first agent searches to optimize token usage.

A conceptual critique of the LLM Wiki pattern examining the boundary between machine bookkeeping and human judgment. Argues that while agents excel at mechanical maintenance (cross-referencing, updating indexes, and linting), compounding knowledge bases require bounded domain scope and active human curation rather than passive all-in-one data dumping.

A conceptual primer and architectural guide breaking down the LLM Wiki pattern. Analyzes the three layers and core operations (ingest, query, lint), contrasts static compile-at-ingest wikis with query-time RAG, examines scaling bottlenecks (index bloat, n-way compilation costs), and details three implementation paths: hand-crafted agent prompts, the CROW classifier-routed engine (
llmwiki2), and no-code desktop applications.
An analysis of Karpathy's LLM Wiki pattern, reviewing top implementations, product principles, and RAG integration.
LLM Wiki: The Compounding Knowledge Base, Explained (Mikko Lehtimäki on Softlandia Blog)softlandia.comA conceptual primer explaining how LLM Wikis function as self-updating knowledge bases for AI agents. Compares link-based traversal against vector RAG models, outlining the minimal operational structure (sources/, wiki/, index.md, and system instructions) needed to enable compounding context.

An architectural walkthrough analyzing the convergence of LLM-maintained wikis across developer tools (OpenWiki, DeepWiki, AutoWiki). Identifies the critical failure mode of knowledge base poisoning and compression drift—where repeated summarization distorts ground-truth nuances—and advocates for a verification trust layer to benchmark provenance.

An article analyzing the lifecycle of personal knowledge management systems (Zettelkasten, PARA, digital gardens). Argues that standard pipelines neglect knowledge decay and outlines how automated agentic verification loops serve as a maintenance layer to keep flat-file notes grounded and accurate.

An announcement of Pinecone Nexus, a knowledge engine that compiles raw data into task-optimized markdown artifacts.
Self-Authoring LLM Knowledge Basesricardodecal.comA technical conceptualization extending the compilation loop to live developer conversations, transforming ephemeral terminal and editor interactions into structured, persistent memory.

A conceptual breakdown outlining the structural advantages of using compiled Markdown wikis to ground AI chatbots, covering setup simplicity, response speed, contextual consistency, and cost metrics compared to vector RAG setups.

An essay exploring the value of personal, queryable LLM Wikis for product designers. Highlights the challenge of design insights scattered across Figma, Notion, and Slack, proposing a markdown context layer to let models synthesize design learnings and onboarding flows across multi-product histories.

A post comparing nine different implementations to extract the four pillars of agent memory: gathering signals, remembering, dreaming & pruning (updating/purging knowledge), and retrieving & speaking.

An essay analyzing the tension between machine compilation and human curation in LLM Wikis. Argues that a knowledge vault should serve as a "biography of understanding" rather than a database, suggesting that humans must retain responsibility for weighing ideas and tracking perspective changes.

An architectural essay detailing dependency tracking and invalidation cascades within Markdown-based agent memory vaults. Explains how frontmatter
depends_on:declarations enable automated sweeps that flag downstream conclusions asneeds_reviewwhen a foundational premise is superseded, contrasting human-in-the-loop review gates against silent automatic overwrites.
The LLM Wiki: Stop Retrieving, Start Compounding (Rohit Ghumare)aiengineeringfromscratch.comA conceptual primer contrasting stateless vector retrieval with compounding compiled markdown wikis. Explains the three-layer architecture (sources, wiki, schema), the Ingest-Query-Lint execution cycle with query back-filing, the compiler mental model (Obsidian as IDE, LLM as programmer), and the historical connection to Vannevar Bush's Memex.
The Real Second Brain: An Autonomous Knowledge Enginevasileioszografos.substack.comAn architectural essay defining the paradigm shift from ephemeral chat interfaces to persistent, agent-driven local wikis. It details how flat-file markdown vaults serve as a deterministic long-term memory layer for autonomous daemons, allowing collaborative, human-in-the-loop knowledge compilation and state management.

An architectural analysis mapping the LLM Wiki pattern to a compiler architecture: raw sources as source code, ingestion as compilation, markdown as intermediate representation, querying as linking, and linting as optimization. Critiques static knowledge assumptions and explores lifecycle mechanics like confidence decay and supersession chains.

An overview by the Mem0 team analyzing the emerging landscape of model-maintained markdown vaults.

An essay examining the evolving landscape of LLM knowledge bases through Niklas Luhmann's Zettelkasten analogy. Highlights industry consensus around the "file over app" philosophy (storing plain text files in Git) and explores emerging frontiers including agent reading rooms, complex knowledge shapes, and automated knowledge compilation pipelines.

A comparative primer examining three architectural responses to Karpathy's LLM Wiki pattern: Google's Open Knowledge Format (OKF) specification, Tencent's layout-aware WeKnora framework, and LangChain's OpenWiki engine. Explores how each manages persistent knowledge, document parsing, and agent memory.

An architectural critique contrasting Karpathy's compiled external wiki pattern with human-authored personal knowledge management. Compares their maintenance loops, arguing that while LLM-authored wikis can aggressively rewrite external syntheses, human second brains require non-destructive maintenance (diffs and auto-linking) to protect personal voice.

A foundational primer defining the architecture and component layers of external LLM knowledge stores. Distinguishes between vector, graph, and context-window models (such as Karpathy's), diagnoses the main reasons enterprise RAG systems fail (lossy source audits, conflicting schemas), and outlines best practices for upstream metadata enrichment.

A high-level primer introducing the LLM Wiki concept, detailing the three-layer architecture, highlighting Farza's personal wiki case study, and suggesting lighter alternatives for solo operators.
What Is an LLM Wiki? A Simple Guide to Karpathy's Personal Knowledge System (Nick Kirtley on 99helpers)99helpers.comAn introductory conceptual guide explaining the core model of LLM Wikis, contrasting their pre-compiled markdown approach with vector-only RAG pipelines, and detailing formatting best practices (one-idea notes, header summaries).
What Is Andrej Karpathy's LLM Knowledge Base Architecture? The Compiler Analogy Explained (MindStudio Blog)mindstudio.aiAn architectural breakdown explaining the compiler analogy for LLM Wikis, comparing raw sources to source code.

A structural guide defining the core filesystem architecture and ingestion loops of the LLM Wiki pattern. Outlines the read/write boundaries between raw source directories and compiled wiki spaces, detailing a 5-step compilation pipeline and contrasting the pre-compiled text routing method against traditional single-pass vector RAG pipelines.

An architectural essay examining the epistemic lifecycle of agent memory and the challenge of handling dead premises. Details a three-stage Markdown memory model (Remember, Trust, Act) grounded in Karpathy's LLM Wiki pattern, incorporating bi-temporal tracking, dependency-driven invalidation cascades, negative knowledge preservation, and the compilation of replicated wiki procedures into executable SKILL.md agent skills.

An architectural essay arguing that inconsistent agent behavior stems from context deficiency rather than a lack of tool integrations. Outlines why grounding models in a compact, human-maintained Markdown wiki and a repeatable skill produces higher-fidelity outputs than chasing expansive tool ecosystems.
Your Personal Brain: A Practical Guide for Leaders, Practitioners, and Enthusiasts (Nufar Gaspar LinkedIn)linkedin.comA management playbook exploring how to establish a "dump-and-housekeep" maintenance cycle for agent memory.
