By mebius-san
Maintains a curated, LLM-navigable semantic wiki over a markdown+code base — summaries, hierarchical topic tags, and glossed See-also links, kept in sync via git-watch and weekly full-scan routines.
Dispatch when a wiki node needs classification (kind=classify) or See-also linking (kind=link), or when a scope's tag values need consolidating (kind=normalize-tags). It applies its result via a deterministic primitive — apply-node for the per-node kinds, retag for normalize-tags (C-hybrid, no collector). The tail flag (default true) gates only the daemon tail after the apply. tail:true (daemon path): reads its job dir (request.json + source/context), then runs build-index, git-commit, dispatch-link. tail:false (/wiki.relink skill path): no job dir — reads the real files named in the dispatch prompt, applies, then stops (the skill owns build-index/commit). Has Bash; writes node content only via apply-node/retag, never by hand.
Dispatch from /wiki.query once, after the seekers have returned validated entry points. Walks the wiki graph from those entry points — following glossed See-also links and on-demand backlinks (grep) — reads only the relevant node bodies, and returns a synthesised answer with source links. Keeps all traversed node bodies out of the caller's context.
Dispatch from /wiki.query, one per configured wiki scope, to pick entry points for a question. Reads ONLY the one scope's topics.md named in the prompt and returns the relevant node paths (verbatim from the index) with their glosses — never traverses, never reads node bodies. Keeps the large topic catalog out of the caller's context.
Associative Q&A over the wiki graph. Thin dispatcher: a per-scope seeker subagent picks entry points from topics.md, a single gatherer subagent traverses See-also and synthesises the answer. The large topic index and traversed node bodies stay in the subagents' contexts, never the main session.
Bootstrap the lazycortex-wiki plugin for the current project (or globally). Creates the template dir, syncs the navigation rule, seeds wiki settings + agent_models + routines + expert entry. Idempotent — safe to re-run.
Audit a wiki scope's integrity: orphan topics, broken See-also links and repo keys, index desync, missing summaries, stale glosses, unknown axes, duplicate branches, broken code <wiki> blocks, and scope overlaps. Read-only by default; applies fixable repairs only after the operator confirms.
Wizard to create or edit a wiki scope in .claude/lazy.settings.json — collects id, path globs, optional exclude_paths, tag_axes, and topics_index. Strict one-question-per-turn via AskUserQuestion.
Daemon-free, in-session relink of one wiki scope. Computes the relink plan (initial / incremental / anchor-lost) via lazycortex-wiki relink-plan, then dispatches the wiki curator as a synchronous subagent in tail:false mode to classify then link each node — the curator applies its own curation via apply-node (C-hybrid, no collector). The skill rebuilds topics.md once between phases, records the new wiki_synced_sha anchor, and makes the single commit under the operator identity. Use when there is no runtime daemon (the plugin must work standalone) or to force an in-session relink.
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
AI tooling plugins for Claude Code — skills, agents, and hooks that add context optimization, change-history recall, security scanning, and permission management to your Claude Code workflows.
Claude Code grows powerful with skills, agents, and hooks — but unmanaged, it also grows noisy, leaky, and forgetful. LazyCortex is an opinionated toolkit for the same issues you keep hitting:
fix and the PR is closed, is a real cost.Each plugin addresses one of these pains without forcing you to adopt the others.
Core skills, agents, and runtime daemon for Claude Code (expert runtime + agent-model routing + settings management)
Ships 25 skills, 6 agents, 2 commands, 10 rules, and 5 hooks.
See claude/lazycortex-core/ for details.
Format-agnostic diagram engine: /lazy-diagram.draw dispatcher + per-format writer agents (mermaid, ascii, more later). Picks kind and format from request context, ships exemplar templates plus an authoring contract, and bundles a fixture-based regression suite.
Ships 4 skills, 2 agents, 1 command, and 1 rule.
Requires: lazycortex-core
See claude/lazycortex-diagram/ for details.
Generic doc-producing experts (interpreter, designer, planner) plus a starter set of domain aspects (claude-plugin, game-dev, dotfiles). Building blocks — compose specialists in lazy.settings.json[experts] with one agent + one or more aspects.
Ships 1 skill, 3 agents, and 1 command.
Requires: lazycortex-core
See claude/lazycortex-experts/ for details.
Ship lazycortex-core runtime metrics to a Prometheus-compatible observer (Grafana Alloy or OpenTelemetry Collector) — vendor-neutral, observer-server-blind, headless-portable.
Ships 4 skills and 1 command.
Requires: lazycortex-core
See claude/lazycortex-observe/ for details.
Obsidian vault bootstrap and configuration management for Claude Code
Ships 7 skills, 1 agent, 1 command, and 2 hooks.
Requires: lazycortex-core
See claude/lazycortex-obsidian/ for details.
Python coding discipline as a plugin: shared rules + reference guidelines + chk/tst checkers + PostToolUse hook + docstring-writer/test-writer agents + canonical file template. Installs once per repo via /lazy-python.install.
Ships 3 skills, 2 agents, 1 command, 3 rules, and 1 hook.
Requires: lazycortex-core
See claude/lazycortex-python/ for details.
Specification and design skills for Claude Code
Ships 22 skills, 2 agents, and 1 command.
Requires: lazycortex-core, lazycortex-diagram
See claude/lazycortex-specs/ for details.
Maintains a curated, LLM-navigable semantic wiki over a markdown+code base — summaries, hierarchical topic tags, and glossed See-also links, kept in sync via git-watch and weekly full-scan routines.
Ships 5 skills, 3 agents, 1 command, and 1 rule.
Requires: lazycortex-core
See claude/lazycortex-wiki/ for details.
lazycortex-core, lazycortex-obsidian)./lazy-core.install, /lazy-python.install, etc. This drops the plugin's rule templates into .claude/rules/ and sets up any log/changelog scaffolding.npx claudepluginhub mebius-san/lazy-cortex --plugin lazycortex-wikiObsidian vault bootstrap and configuration management for Claude Code
Specification and design skills for Claude Code
Generic doc-producing experts (interpreter, designer, planner) plus a starter set of domain aspects (claude-plugin, game-dev, dotfiles). Building blocks — compose specialists in lazy.settings.json[experts] with one agent + one or more aspects.
Core skills, agents, and runtime daemon for Claude Code (expert runtime + agent-model routing + settings management)
Format-agnostic diagram engine: /lazy-diagram.draw dispatcher + per-format writer agents (mermaid, ascii, more later). Picks kind and format from request context, ships exemplar templates plus an authoring contract, and bundles a fixture-based regression suite.
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
AI-powered wiki generator for code repositories. Generates comprehensive, Mermaid-rich documentation with dark-mode VitePress sites, onboarding guides, deep research, and source citations. Inspired by OpenDeepWiki and deepwiki-open.
Claude + Obsidian knowledge companion. Sets up a persistent, compounding wiki vault (Karpathy's LLM Wiki pattern). v1.7 "Compound Vault" + v1.8 methodology modes close 5 of 5 priority gaps from the May 2026 compass artifact. Ships: substrate alignment with kepano/obsidian-skills, default Obsidian CLI transport, hybrid retrieval (contextual prefix + BM25 + cosine rerank per Anthropic's Sept 2024 research), per-file advisory locking for multi-writer safety, pre-commit verifier agent, AND methodology modes (LYT / PARA / Zettelkasten / Generic) for first-class organizational support no other Claude+Obsidian competitor offers. v1.7.x audit closure: every BLOCKER + HIGH + MEDIUM + LOW finding from the v1.7.0 audit is CLOSED or DEFERRED-with-rationale. Optional DragonScale Memory extension (log folds, deterministic addresses, semantic tiling lint, boundary-first autoresearch).
Complete AI coding workflow system. Self-correcting memory + persistent FTS5-indexed research wikis + auto-research loop + multi-LLM council on a single SQLite store. 33 skills, 8 agents, 22 commands, 37 hook scripts across 24 events. Cross-agent via SkillKit.
Make your AI agent code with your project's architecture, rules, and decisions.
Comprehensive C4 architecture documentation workflow with bottom-up code analysis, component synthesis, container mapping, and context diagram generation