maria/setup-llm-wiki
Bootstrap a new LLM-maintained wiki at a chosen folder, following the llm-wiki.md pattern (a three-layer memex - raw sources, LLM-generated wiki pages, and a CLAUDE.md or AGENTS.md schema that tells the LLM how to ingest and maintain the wiki). Creates the directory layout, writes a tailored schema file, plus index.md and log.md with a bootstrap entry. Use this skill when the user asks to "set up an llm-wiki", "create an LLM wiki", "bootstrap a wiki", "instantiate the llm-wiki pattern", or invokes /setup-llm-wiki. The skill asks the user about target folder, domain (research deep-dive / personalised work wiki / personal knowledge base / business-team wiki / reading a book / combination), source types (web articles, academic PDFs, meeting/podcast transcripts, own notes), image handling, optional search tooling (qmd), schema filename (CLAUDE.md or AGENTS.md), and optional symlinks to sibling folders if the target sits inside an Obsidian vault.
87
90%
Does it follow best practices?
Impact
83%
2.24xAverage score across 4 eval scenarios
Advisory
Suggest reviewing before use
Quality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is an excellent skill description that thoroughly covers what the skill does, when to use it, and provides rich detail about the interactive configuration process. It includes explicit trigger phrases, a clear 'Use this skill when...' clause, and highly distinctive terminology that minimizes conflict risk. The description uses proper third-person voice throughout and balances comprehensiveness with clarity.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description lists multiple specific concrete actions: creates directory layout, writes a tailored schema file, creates index.md and log.md with a bootstrap entry. It also details the three-layer memex architecture and enumerates the specific configuration questions asked of the user. | 3 / 3 |
Completeness | Clearly answers both 'what' (bootstraps a three-layer memex wiki with directory layout, schema file, index.md, and log.md) and 'when' (explicit 'Use this skill when...' clause with multiple trigger phrases). The description also details the interactive configuration steps. | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural trigger terms including 'set up an llm-wiki', 'create an LLM wiki', 'bootstrap a wiki', 'instantiate the llm-wiki pattern', and the slash command '/setup-llm-wiki'. These cover multiple natural phrasings a user might employ. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive with a very specific niche: bootstrapping an LLM-maintained wiki following a particular three-layer memex pattern. The specific terminology ('llm-wiki', 'memex', 'CLAUDE.md', 'AGENTS.md') and the slash command make it unlikely to conflict with other skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-crafted, highly actionable skill with excellent workflow clarity and concrete executable guidance. Its main weakness is that it's somewhat verbose for a skill file — the extensive placeholder documentation and template content blocks inflate the token cost, and the referenced bundle files (templates, scripts, fragments) are not provided, making it impossible to verify the progressive disclosure structure actually works. The skill does an excellent job of handling edge cases (existing files, vault detection, conditional sections) with explicit validation steps.
Suggestions
Move the detailed placeholder replacement content (the full markdown blocks for each {{PLACEHOLDER}}) into the actual template files they reference, keeping only brief descriptions in SKILL.md to reduce token footprint.
Provide the referenced bundle files (assets/CLAUDE.md.template, scripts/create_dirs.sh, assets/fragments/*.md, references/obsidian-graph.md) so the progressive disclosure structure is complete and verifiable.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is fairly detailed and well-structured, but includes some verbose sections that could be tightened — e.g., the full reference phrasings for each domain type, the exhaustive placeholder documentation, and the conditional fragment explanations. However, most content is genuinely instructive and not explaining things Claude already knows. | 2 / 3 |
Actionability | The skill provides highly concrete, executable guidance: specific script commands (`scripts/create_dirs.sh <target> <source_types_csv>`), exact placeholder names and their replacement content, shell commands for symlinks and vault detection, exact file paths, and precise template content blocks. Everything is copy-paste ready or directly implementable. | 3 / 3 |
Workflow Clarity | The workflow is clearly sequenced in 8 numbered steps with explicit validation checkpoints — checking for non-empty target folders, never overwriting existing files without confirmation, conditional symlink creation, and a final summary step. The input-gathering phase is also well-sequenced with clear dependencies (e.g., symlinks are conditional on Obsidian vault detection). | 3 / 3 |
Progressive Disclosure | The skill references multiple template files and fragment files (e.g., `assets/CLAUDE.md.template`, `assets/fragments/image-handling.md`, `references/obsidian-graph.md`, `scripts/create_dirs.sh`) which is good progressive disclosure design, but no bundle files are provided. The SKILL.md itself is quite long (~200 lines) with inline template content that could arguably live in the referenced asset files, and the placeholder documentation section is essentially inlined reference material. | 2 / 3 |
Total | 10 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
Reviewed
Table of Contents