llm-wiki-bootstrap
Bootstrap and operate an LLM Wiki: a persistent Markdown knowledge base with immutable raw sources, an LLM-maintained wiki, a SCHEMA.md operating contract, thin runtime pointer files, EXTEND.md preferences, and optional local BM25 search. Use when the user asks to create or bootstrap an LLM wiki, ingest sources, query or lint a wiki, configure EXTEND.md preferences, or set up BM25/full-text search for an LLM Wiki.
What this skill does
# LLM Wiki Bootstrap Build and maintain a personal LLM Wiki: raw evidence stays immutable in `raw/`, durable knowledge is compiled into `wiki/`, and `SCHEMA.md` is the generated wiki's single operating contract. This file is intentionally short. Load detailed references only for the operation the user requested. ## Operating Model - Keep `SCHEMA.md` as the single source of truth inside every generated wiki. - Keep `wiki/concept-table.md` as the maintained concept map: it complements `wiki/index.md` by summarizing durable concepts, relationships, sources, status, and maintenance notes. - Keep `CLAUDE.md`, `AGENTS.md`, and `.github/copilot-instructions.md` as thin pointers to `SCHEMA.md`; never duplicate operating rules into them. - Treat `raw/` as read-only evidence. Create and update derived knowledge only under `wiki/`. - Apply `EXTEND.md` preferences before bootstrap, ingest, query, lint, or BM25 work. - Ask only for missing decisions, and batch setup questions into one user prompt when possible. - Prefer progressive disclosure: read the smallest reference file that can complete the current task. ## Intent Router | User intent | Do this | Reference | | -------------------------------------- | -------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------- | | Create or bootstrap a new wiki | Run preference preflight, gather setup answers, then generate the scaffold and seed files. | `references/workflows/bootstrap.md` | | Ingest one or more sources | Run preference preflight, check the optional BM25 gate, then compile source knowledge into `wiki/`. | `references/workflows/ingest.md` | | Answer a domain question from the wiki | Run preference preflight, navigate `wiki/index.md`, use BM25 only as a candidate finder, then cite wiki pages. | `references/workflows/query.md` | | Check wiki health or consistency | Run preference preflight, scan structure and content, report findings, then fix only approved items. | `references/workflows/lint.md` | | Configure or use BM25 search | Load preferences, ask before initialization, create local search files, smoke test, and log results. | `references/workflows/bm25.md` | | Configure preferences | Locate or create `EXTEND.md` using the schema and template. | `references/config/extend-schema.md`, `references/templates/extend.md` | ## Preference Preflight Before bootstrap, ingest, query, lint, or BM25 work, read `references/config/extend-schema.md` and use the first available `EXTEND.md` in its lookup order. If none exists, run the first-time preference setup and create one from `references/templates/extend.md` before continuing. On first use in a session, briefly say which preference file is active. Do not silently use defaults. ## Reference Loading Rules - For bootstrap, read `references/workflows/bootstrap.md`, then load each template only when creating that file. - For ingest, query, and lint, read the matching workflow file first; read `references/workflows/bm25.md` only when preferences or the user request require search behavior. - For schema generation, read `references/templates/schema.md` and inject domain snippets from `references/templates/domain-page-types.md` as needed. - For pointer files, read `references/templates/agent-pointer.md` and keep the generated file minimal. - For BM25 initialization, read `references/templates/wiki_fts.py` and `references/templates/bm25-readme.md` only after the user approves initialization. ## Package Layout Rules - Put skill-owned runnable helpers in `scripts/` if they are meant to be executed from the installed skill package. - Put generated-output templates in `references/templates/`, even when the generated file will become executable after it is copied into the user's wiki. - Put long procedural documentation in `references/workflows/` and configuration schemas in `references/config/`. - In this skill, `references/templates/wiki_fts.py` is a template for the generated wiki's `scripts/wiki_fts.py`; it is not executed in place by the skill. ## Hard Rules - Never overwrite an existing wiki without explicit user approval. - Never modify files in `raw/` except to create empty scaffold placeholders such as `.gitkeep` during bootstrap. - Never answer directly from BM25 snippets; open and read the returned wiki pages first. - Preserve source language during ingest unless the user asks for a different wiki language. - Update `wiki/index.md` and append `wiki/log.md` whenever wiki content changes. - If a search index rebuild fails, keep wiki edits, log the failure, and fall back to `wiki/index.md` plus text search. ## Core Outputs A successful bootstrap creates a wiki root containing: ```text raw/ wiki/index.md wiki/concept-table.md wiki/log.md wiki/overview.md SCHEMA.md .gitignore ``` Optional outputs depend on setup answers and preferences: runtime pointer files, `.vscode/settings.json`, `scripts/wiki_fts.py`, `indexes/`, and `exports/`.
Related in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.