llmwiki

A persistent, file-based memory layer for coding agents — give Claude Code, Codex & others a project brain (durable decisions, requirements, constraints) via a zero-dependency CLI.

A local-first personal knowledge layer for AI agents. Build evolving knowledge spaces with LLM Wiki, knowledge graphs, and agent workflows.

Open-source, local-first AI agent for coding and real work. BYOK models, MCP, skills, plugins, workflows, and private knowledge bases.

The one data index for your agentic applications

LLMWikiNG is a local, privacy-friendly wiki platform developed by ZeroDot1 and based on the Karpathy LLM Wiki Pattern. The wiki is maintained and expanded by AI assistants—and you can easily read, search, and manage it right in your browser.

开源 · 本地优先的 AI 智能体：编程与实干兼备。 自带模型密钥，内置 MCP、技能、插件、工作流、知识库以及数据库。

llmwiki (Quiz_wiki) — auto-captures your coding-agent sessions (Claude Code / Codex / OpenCode) into a compounding, local-first project wiki that quizzes you back. The llmwiki quiz is a daily forgetting-curve quiz (/wiki-quiz) that keeps your memory of your own decisions sharp. Markdown as source of truth, no MCP, no build.

The intelligence layer behind the organization

A practical playbook for building a personal knowledge system with Obsidian, Claude, Codex, and an LLMWiki-style folder structure.

JanusLM is a Two-faced AI agent. One side reads, catalogs, connects. The other acts: writes documents, analyzes data, answers questions.

Access your markdown vault via ChatGPT/Claude on Mobile. Obsidian sync compatible.

Serve local Markdown/wiki knowledge folders as agent-readable HTTP, MCP-style, and A2A-style context.

Agent-native LLM Wiki (Karpathy-style) RAG: distill raw docs → structured Markdown wiki → hybrid search via MCP

Local-first knowledge app that turns documents, images, and web pages into a connected, searchable, Obsidian-compatible Markdown wiki.

Static-site template for cross-linked markdown wikis. Themable, search + knowledge graph, deploys to GitHub Pages.

Structural LLM Wiki (sllmwiki) is an agent skill, packaged as a Claude Code plugin, that turns a git repository into a wiki your coding agent researches and keeps current.

Make small AI models reliable at editing Markdown. Incise provides safe, semantic, byte-preserving edits for AI agents.

Weave Skills provides small filesystem-native cognitive operational overlays that keep reasoning operational across long-running engineering work, enabling implementation-grounded continuity and contextual recovery across AI-assisted development sessions.

Agent-native LLM Wiki (Karpathy-style) RAG: distill raw docs → structured Markdown wiki → hybrid search via MCP

Source-backed memory for long-running AI agents: maintained synthesis, freshness review, evidence links, and MCP/API access.