--- title: So I tried Matt's skills... | daily.dev description: A deep dive into two popular collections of AI agent 'skills' (markdown-based instruction files): Matt Pocock's skills repo (200k+ GitHub stars) and Lauren... canonical: https://daily.dev/posts/so-i-tried-matt-s-skills--0f5yy5jlx twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: So I tried Matt's skills... | daily.dev og:description: A deep dive into two popular collections of AI agent 'skills' (markdown-based instruction files): Matt Pocock's skills repo (200k+ GitHub stars) and Lauren... og:url: https://daily.dev/posts/so-i-tried-matt-s-skills--0f5yy5jlx og:image: https://api.daily.dev/og/posts/0f5yY5Jlx.png og:image:alt: So I tried Matt's skills... og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # So I tried Matt's skills... **[Theo - t3․gg](https://daily.dev/sources/t3dotgg)** · 38 min read · 29 upvotes · 11 comments ## Summary A deep dive into two popular collections of AI agent 'skills' (markdown-based instruction files): Matt Pocock's skills repo (200k+ GitHub stars) and Lauren 'Potato' Tan's Pstack skills. The reviewer tests standout skills like unslop (removes AI writing tells and adds human voice), grilling/grill-me (relentless interview-style planning), blast-radius (impact analysis before changes), show-your-work, wizard, writing-for-agents, and wait-what. Includes live before/after comparisons showing how unslop dramatically improves AI-generated text readability, and a workflow for auditing personal usage against available skills to decide which to adopt rather than blindly copying setups. Sponsored segment for Depot's CI/build caching and remote agent sandboxes. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What does the unslop AI agent skill actually do? It rewrites AI-generated text to remove robotic patterns and add a human voice. The process scans for tells like corporate puffery ('pivotal moment', 'testament to'), overused em-dashes, chatbot phrases ('I hope this helps'), and vague jargon, then rewrites while preserving meaning, adding opinion and varied sentence rhythm, and self-auditing by asking what still reads as AI-generated. _Developers tuning their own agent prompts for less robotic output can find more workflow breakdowns like this on daily.dev._ ### What is the difference between Matt Pocock's skills repo and Potato's Pstack skills for AI coding agents? Matt Pocock's repo has over 200,000 GitHub stars and includes skills like grill with docs, diagnosing bugs, wizard, and writing for agents, often built around a prescriptive workflow from planning to code review. Pstack, created by former React core team member Lauren 'Potato' Tan (now at Cursor), includes skills like unslop, arena, blast radius, and show-your-work, generally considered more readable but somewhat cursor-specific. _Comparing agent skill ecosystems like these helps developers pick a setup on daily.dev before committing to one workflow._ ### How can I test an AI agent skill without formally installing it? Copy the skill's markdown text directly and paste it into a prompt to your coding agent along with your request; the agent will follow the instructions as if the skill were installed, since skills are just markdown files read into context. This works for text-only skills, though skills bundling scripts require pointing the agent to fetch or download those separately. _Trying out agent skills before committing to a setup is easier when developers share findings, as they do on daily.dev._ ## Community discussion Top comments from developers on daily.dev. **@confidentcoding** · 4 upvotes > Huh, I've never really thought of using anybody else's skills, I always just write my own, since I want them to be highly contextual to my own codebase, tools, and workflows (and it's really not that hard; we're just talking text files, here). If I come across any skill that seems interesting, I just replicate it inside one of my own projects. Zed has a fantastic feature to easily create them, too. **@ilijaveselinovic** · 3 upvotes > Matt's skills are great. But it's even better when you customize the existing skills for your needs. Then it's a premium. It's not that easy, but whoever succeeds in it gets even better results. **@hariomlohardev** · 3 upvotes > One of the best skills for claude code > >  **@fabricioereche** · 1 upvotes > Summary of the video: Mat skills are good (5 min of video)…. But did you took a look into Lauren skills?! (All the rest of the video) **@ivanoiorio** · 0 upvotes > Two reasons not to use it: > > 1) You have to trust that it isn't a waste of tokens; why not simply put "use TDD methodology" in `claude.md`? You have to do that anyway, and Claude _already knows_ what that entails without needing it explained all over again. > > 2) The worst part: you don't know if it's actually right for your needs; it's often verbose, and Claude writes in an incomprehensible way, leaving you arguing about absurd things. > > 3) It obscures Claude Code's native "planning" capability, which could improve in lockstep with Anthropic's models... I mean... can it really compete with... ## Similar posts on daily.dev - [thoughtbot's agent skills](https://daily.dev/posts/thoughtbot-s-agent-skills-gpmz89erh) · thoughbot · 12 upvotes · 2 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#prompt-engineering](https://daily.dev/tags/prompt-engineering), [#claude-code](https://daily.dev/tags/claude-code) [View this post on daily.dev](https://daily.dev/posts/so-i-tried-matt-s-skills--0f5yy5jlx) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"So I tried Matt's skills...","url":"https://daily.dev/posts/so-i-tried-matt-s-skills--0f5yy5jlx","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/so-i-tried-matt-s-skills--0f5yy5jlx"},"datePublished":"2026-08-19T02:35:40.114Z","dateModified":"2026-09-13T21:35:00.005Z","description":"A deep dive into two popular collections of AI agent 'skills' (markdown-based instruction files): Matt Pocock's skills repo (200k+ GitHub stars) and Lauren...","image":"https://i.ytimg.com/vi/0oXOOlqVu5M/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/0oXOOlqVu5M/sddefault.jpg","isAccessibleForFree":true,"articleSection":"Theo - t3․gg","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Theo - t3․gg","logo":"https://media.daily.dev/image/upload/s--UmX7IyU3--/f_auto/v1704628081/logos/t3dotgg.jpg","url":"https://daily.dev/sources/t3dotgg"},"commentCount":11,"discussionUrl":"https://daily.dev/posts/so-i-tried-matt-s-skills--0f5yy5jlx","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":29},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":11}],"keywords":"ai-agents,prompt-engineering,claude-code","timeRequired":"PT38M","video":{"@type":"VideoObject","name":"So I tried Matt's skills...","description":"A deep dive into two popular collections of AI agent 'skills' (markdown-based instruction files): Matt Pocock's skills repo (200k+ GitHub stars) and Lauren...","thumbnailUrl":"https://i.ytimg.com/vi/0oXOOlqVu5M/sddefault.jpg","uploadDate":"2026-08-19T02:35:40.114Z","duration":"PT38M","url":"https://api.daily.dev/r/0f5yY5Jlx","embedUrl":"https://www.youtube.com/embed/0oXOOlqVu5M"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Theo - t3․gg","item":"https://daily.dev/sources/t3dotgg"},{"@type":"ListItem","position":3,"name":"So I tried Matt's skills..."}]} {"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/so-i-tried-matt-s-skills--0f5yy5jlx","comment":[{"@type":"Comment","text":"Huh, I’ve never really thought of using anybody else’s skills, I always just write my own, since I want them to be highly contextual to my own codebase, tools, and workflows (and it’s really not that hard; we’re just talking text files, here). If I come across any skill that seems interesting, I just replicate it inside one of my own projects. Zed has a fantastic feature to easily create them, too.","datePublished":"2026-08-20T17:30:07.597Z","dateModified":"2026-08-20T17:30:29.777Z","url":"https://daily.dev/posts/0f5yY5Jlx#c-qqQHJN3IV","author":{"@type":"Person","name":"Lars Faye | Confident Coding","url":"https://daily.dev/confidentcoding","image":"https://media.daily.dev/image/upload/s--OGZu5DEc--/f_auto/v1772569630/avatars/avatar_umWZ9aQAng34qk5aaJl2q?_a=BAMAMiiu0"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":4}},{"@type":"Comment","text":"Matt’s skills are great. But it’s even better when you customize the existing skills for your needs. Then it’s a premium. It’s not that easy, but whoever succeeds in it gets even better results.","datePublished":"2026-08-19T05:51:14.867Z","url":"https://daily.dev/posts/0f5yY5Jlx#c-32wrH4wQ3","author":{"@type":"Person","name":"Ilija","url":"https://daily.dev/ilijaveselinovic","image":"https://media.daily.dev/image/upload/s--GqTPoTzA--/f_auto/v1738848590/avatars/avatar_Nes9qpxMjBg8ZBR90Du7C"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":3}},{"@type":"Comment","text":"One of the best skills for claude code","datePublished":"2026-08-20T03:10:14.520Z","url":"https://daily.dev/posts/0f5yY5Jlx#c-Vnqh2tWr0","author":{"@type":"Person","name":"Hariom","url":"https://daily.dev/hariomlohardev","image":"https://avatars.githubusercontent.com/u/264109575?v=4"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":3}},{"@type":"Comment","text":"Summary of the video: Mat skills are good (5 min of video)…. But did you took a look into Lauren skills?! (All the rest of the video)","datePublished":"2026-08-23T19:13:49.857Z","url":"https://daily.dev/posts/0f5yY5Jlx#c-bUMbuvAsO","author":{"@type":"Person","name":"Fabricio Reche","url":"https://daily.dev/fabricioereche","image":"https://media.daily.dev/image/upload/s--6WAbm-qe--/f_auto/v1736291520/avatars/avatar_Q1tbYX25I03fRdERtm6Rr"},"interactionStatistic":{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1}},{"@type":"Comment","text":"Two reasons not to use it:\n\n\nYou have to trust that it isn’t a waste of tokens; why not simply put “use TDD methodology” in claude.md? You have to do that anyway, and Claude already knows what that entails without needing it explained all over again.\n\n\nThe worst part: you don’t know if it’s actually right for your needs; it’s often verbose, and Claude writes in an incomprehensible way, leaving you arguing about absurd things.\n\n\nIt obscures Claude Code’s native “planning” capability, which could improve in lockstep with Anthropic’s models… I mean… can it really compete with that?\n\n\nI’ve tried it, and for me, it’s something to be ABSOLUTELY avoided.","datePublished":"2026-09-09T13:05:10.504Z","url":"https://daily.dev/posts/0f5yY5Jlx#c-51z16oCeb","author":{"@type":"Person","name":"Ivano Iorio","url":"https://daily.dev/ivanoiorio","image":"https://lh3.googleusercontent.com/a/ACg8ocKA4wD4lM9V5uEQ17xEAUPziv77OEdGj04wZ-uZPG5H9l7CEukn=s96-c"}}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/so-i-tried-matt-s-skills--0f5yy5jlx#faq","mainEntity":[{"@type":"Question","name":"What does the unslop AI agent skill actually do?","acceptedAnswer":{"@type":"Answer","text":"It rewrites AI-generated text to remove robotic patterns and add a human voice. The process scans for tells like corporate puffery ('pivotal moment', 'testament to'), overused em-dashes, chatbot phrases ('I hope this helps'), and vague jargon, then rewrites while preserving meaning, adding opinion and varied sentence rhythm, and self-auditing by asking what still reads as AI-generated. Developers tuning their own agent prompts for less robotic output can find more workflow breakdowns like this on daily.dev."}},{"@type":"Question","name":"What is the difference between Matt Pocock's skills repo and Potato's Pstack skills for AI coding agents?","acceptedAnswer":{"@type":"Answer","text":"Matt Pocock's repo has over 200,000 GitHub stars and includes skills like grill with docs, diagnosing bugs, wizard, and writing for agents, often built around a prescriptive workflow from planning to code review. Pstack, created by former React core team member Lauren 'Potato' Tan (now at Cursor), includes skills like unslop, arena, blast radius, and show-your-work, generally considered more readable but somewhat cursor-specific. Comparing agent skill ecosystems like these helps developers pick a setup on daily.dev before committing to one workflow."}},{"@type":"Question","name":"How can I test an AI agent skill without formally installing it?","acceptedAnswer":{"@type":"Answer","text":"Copy the skill's markdown text directly and paste it into a prompt to your coding agent along with your request; the agent will follow the instructions as if the skill were installed, since skills are just markdown files read into context. This works for text-only skills, though skills bundling scripts require pointing the agent to fetch or download those separately. Trying out agent skills before committing to a setup is easier when developers share findings, as they do on daily.dev."}}]} ```