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 blader · Claude Skill · ★ 617
Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h
🔒 Is napkin safe to install? View the security audit →
Napkin A skill for Claude Code that gives the agent persistent memory of its mistakes. The agent maintains a markdown file in your repo () where it tracks what went wrong, what you corrected, and what worked. It reads the file at session start and writes to it continuously as it works. By session 3-5 the behavior shift is significant — the agent stops making mistakes you've already corrected and starts pre-empting issues before you catch them. Baby continual learning in a markdown file. Install Claude Code Codex That's it. The skill activates every session, unconditionally. How it works Session start: Agent reads in the current repo. If it doesn't exist, it creates one. During work: Agent logs mistakes (its own and yours), corrections, patterns, preferences — as they happen, not just at session end. Over sessions: The napkin compounds. Session 1 is normal. Session 3 the agent is catching things before you do. Session 5 it's a different tool.
| Stars | 617 |
| Forks | 48 |
| Category | Claude Skill |
| License | MIT |
| Quality Score | 57.810117593954/100 |
| Open Issues | 1 |
| Last Updated | 2026-02-21 |
| Created | 2026-02-07 |
| Platforms | claude-code |
| Est. Tokens | ~0k |
These tools work well together with napkin for enhanced workflows:
Looking for a napkin alternative? If you're comparing napkin with other claude skill tools, these 4 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.
Battle-tested Claude Code, OpenAI Codex, Cursor configs, plugins, hooks and agents with Kimi, MiniMax and GLM
An open-source AI marketing operating system for strategy, SEO, AEO/GEO, paid media, content, CRM, and analyti
Repo for the Claude Code Marketplace to use with the Claude for Life Sciences Launch. This will continue to ho
Report issues related to MCP integration with Claude here.
Explore other popular claude skill tools:
napkin is A Claude Code skill that gives the agent persistent memory of its mistakes via a per-repo markdown scratchpad.. It is categorized as a Claude Skill with 617 GitHub stars.
You can find installation instructions and usage details in the napkin GitHub repository at github.com/blader/napkin. The project has 617 stars and 48 forks, indicating an active community.
napkin is released under the MIT license, making it free to use and modify according to the license terms.
The top alternatives to napkin on Agent Skills Hub include claude-codex-settings, digital-marketing-pro, life-sciences. 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: