No red flags found in any of the 11 categories — no credential harvesting, no data exfiltration, no curl-pipe-shell installer. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →
by toolboxmd · Claude Skill · ★ 103
Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h
🔒 Is karpathy-wiki safe to install? View the security audit →
Karpathy Wiki A Claude Code skill for auto-maintained LLM wikis — based on Andrej Karpathy's LLM Wiki pattern. Instead of re-deriving answers from raw documents every time (RAG), the LLM incrementally builds and maintains a wiki — a structured, interlinked collection of markdown files. The wiki compounds with every source you add and every question you ask. For day-to-day usage and the workflow walkthrough, see MANUAL.md. What it does As you work with Claude Code, any durable knowledge — research findings, resolved confusions, validated patterns, gotchas, architectural decisions — gets written as a small capture file and processed by a detached background worker into a persistent wiki. The wiki is git-versioned. Your flow is never interrupted. User commands: — health report (categories, page counts, depth-violation count, soft-ceiling indicator, quality rollup, last ingest, drift, recent ingester-reported issues, fork-asymmetry). Implemented. — write a chat-driven capture (the agent's canonical entry point; supports , body via stdin or ). — drift-scan + drain on demand. — show recent ingester-reported issues, grouped and severity-ordered. — change per-cwd wiki mode.
| Stars | 103 |
| Forks | 25 |
| Language | Shell |
| Category | Claude Skill |
| License | MIT |
| Quality Score | 66.9632152900827/100 |
| Open Issues | 6 |
| Last Updated | 2026-08-27 |
| Created | 2026-04-04 |
| Platforms | claude-code, cli |
| Est. Tokens | ~13k |
Looking for a karpathy-wiki alternative? If you're comparing karpathy-wiki with other claude skill tools, these 2 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.
Report issues related to MCP integration with Claude here.
A tool for autonomous agents like OpenClaw to discover and learn skills autonomously
Explore other popular claude skill tools:
karpathy-wiki is Karpathy Wiki - Claude Code skills for building persistent, compounding knowledge bases. Based on Andrej Karpathy's LLM Wiki pattern.. It is categorized as a Claude Skill with 103 GitHub stars.
karpathy-wiki is primarily written in Shell.
You can find installation instructions and usage details in the karpathy-wiki GitHub repository at github.com/toolboxmd/karpathy-wiki. The project has 103 stars and 25 forks, indicating an active community.
karpathy-wiki is released under the MIT license, making it free to use and modify according to the license terms.
The top alternatives to karpathy-wiki on Agent Skills Hub include claude-ai-mcp, skyll. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.
Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.
The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.
Sources & who's responsible: