llm-wiki
Compiles and maintains a persistent LLM-written markdown wiki between immutable raw sources and answers—the Karpathy LLM Knowledge Base pattern. The agent writes and maintains the wiki; the human curates sources and reads it. Knowledge compounds instead of being re-derived each query. Triggers: llm wiki, persistent wiki, personal knowledge base, wiki maintenance, ingest sources, compound knowledge, index.md, log.md, Obsidian wiki, cross-references, Karpathy wiki pattern, compile wiki, knowledge base. Uses: Read, Glob, Grep, file edits, optional WebSearch, AskUserQuestion when schema or goals are ambiguous. Outputs: updated wiki pages, index, append-only log, citations on query, filed answers, visual outputs. Do NOT use for: mutating raw sources, one-off chat answers with no wiki artifact, or replacing a user-defined wiki schema without reading it first.
94
92%
Does it follow best practices?
Impact
96%
0.97xAverage score across 3 eval scenarios
Passed
No known issues
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 a strong, well-crafted skill description that excels across all dimensions. It clearly defines a specific niche (persistent LLM-maintained wiki using the Karpathy pattern), provides extensive trigger terms, explicitly lists tools and outputs, and includes negative triggers to reduce false matches. The description is detailed without being padded, and uses appropriate third-person voice throughout.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: 'writes and maintains the wiki', 'curates sources', 'ingest sources', 'compile wiki', 'updated wiki pages, index, append-only log, citations on query, filed answers, visual outputs'. Also specifies tools used and explicit exclusions. | 3 / 3 |
Completeness | Clearly answers 'what' (compiles and maintains a persistent LLM-written markdown wiki, writes wiki pages, index, log, citations) and 'when' (explicit Triggers list plus 'Do NOT use for' negative triggers). Both dimensions are thoroughly addressed. | 3 / 3 |
Trigger Term Quality | Provides extensive natural trigger terms including 'llm wiki', 'persistent wiki', 'personal knowledge base', 'wiki maintenance', 'ingest sources', 'compound knowledge', 'index.md', 'log.md', 'Obsidian wiki', 'Karpathy wiki pattern', 'compile wiki', 'knowledge base'. These cover many natural variations a user might say. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive with a clear niche: the 'Karpathy LLM Knowledge Base pattern' is a very specific concept. The combination of persistent wiki maintenance, append-only logs, immutable raw sources, and compound knowledge creates a unique profile unlikely to conflict with generic note-taking or document skills. The 'Do NOT use for' section further sharpens boundaries. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
85%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a strong, well-structured skill that clearly defines a novel workflow pattern with concrete processes, explicit constraints, and good examples. Its main weakness is moderate verbosity—some concepts are stated in non-negotiables, restated in process steps, and restated again in anti-patterns. The progressive disclosure and workflow clarity are excellent, with appropriate use of companion files and explicit validation steps throughout.
Suggestions
Consolidate the anti-patterns section by removing items that directly restate non-negotiables (e.g., 'Editing raw sources' and 'Skipping index or log' are already covered above) to reduce redundancy and save tokens.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is well-written but somewhat verbose for what it conveys. The integrated examples are valuable but could be tighter—some bullet points restate what was already covered in the process sections. The 'three layers' table and 'scaling' section add moderate value but the anti-patterns section partially restates the non-negotiables. | 2 / 3 |
Actionability | The skill provides highly concrete, step-by-step processes for all three operations (ingest, query, lint) with specific file paths, exact log heading formats, and two detailed worked examples showing exact agent actions. The log format is precisely specified with explicit keyword constraints. | 3 / 3 |
Workflow Clarity | All three workflows (ingest, query, lint) are clearly sequenced with numbered steps. Validation checkpoints are present: read schema first, read index first on query, flag contradictions explicitly, and the lint process itself serves as a validation/feedback loop. The 'non-negotiables' section adds explicit guardrails for destructive operations (never modify raw sources). | 3 / 3 |
Progressive Disclosure | The skill has a clear overview structure with well-signaled one-level-deep references to companion rules (index-log-conventions.md, optional-tooling.md). Detailed conventions are appropriately split into separate files rather than inlined. The main document stays focused on behaviors and process. | 3 / 3 |
Total | 11 / 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.
- Repository
- kvokov/oh-my-ai
- Commit
75611cd
- Reviewed
Table of Contents
Is this your skill?
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.