正在加载视频...

视频加载失败

Introducing Skill Manager. Your coding agent loads every skill's description, whether it's needed or not wasting precious context and potentially loading conflicting skills. With Skill Manager you can switch skills on and off one at a time, group them by theme either manually or by using the AI auto-group...

21,067 次观看 • 6 天前 •via X (Twitter)

27 条评论

Shahid🇵🇸 的头像
Shahid🇵🇸6 天前

for claude code, does off mean moving the skill out of the folder? there's a middle state worth a toggle: disable-model-invocation keeps the description out of context but /name still works, and skillOverrides in settings does the same for skills you didn't write without touching their files

camsoft2000 的头像
camsoft20006 天前

It moves the skill folder to a holding directory. The middle state is an interesting idea, I'll explore it! Thanks.

eric provencher 的头像
eric provencher6 天前

This is very needed

camsoft2000 的头像
camsoft20006 天前

Indeed!

Zannatun Sheum 的头像
Zannatun Sheum6 天前

Context tax from unused skill descriptions is real. Being able to toggle them per session is the missing piece.

Mahdi 的头像
Mahdi6 天前

This is the bit that bites me. Half the skills never fire and they still burn context every turn.

Anderson 的头像
Anderson6 天前

built a manager for the skills your agent cant manage on its own, we are three layers deep now

camsoft2000 的头像
camsoft20006 天前

Sadly. Yes.

Anderson 的头像
Anderson6 天前

turtles all the way down

Artem Novichkov 的头像
Artem Novichkov6 天前

Awesome domain 😅

camsoft2000 的头像
camsoft20006 天前

Ah you noticed! 🤣

Vedran Burojević 的头像
Vedran Burojević6 天前

Nice one! Thanks for sharing!

MoNagm 的头像
MoNagm6 天前

Context bloat from skills you aren't using is so real. Does it support profiles? I'd want a writing set and a coding set I can flip between.

camsoft2000 的头像
camsoft20006 天前

Yes! Though I call them groups.

Ben Mo 的头像
Ben Mo6 天前

do group changes take effect in an already running agent session, or does it need a fresh session?

camsoft2000 的头像
camsoft20006 天前

Depends on the client, Codex automatically detects skill changes in real-time but you'll most likely need to start a new session, which is best practice when switching tasks anyway.

catman 的头像
catman6 天前

It’s like keeping only the tools you need on the workbench: toggling skills reduces clutter in the agent’s context and makes conflicts easier to spot.

Bharath 的头像
Bharath6 天前

Love the vibe colors of the website. Congratulations on the launch!!

camsoft2000 的头像
camsoft20006 天前

Thanks, finally launched something. Surprisingly it was Claude Code that got me over the line, Codex just wasn't cutting it.

Bharath 的头像
Bharath6 天前

I would love to hear any skills used for the design / UX. I am wandering and trying out but found none to be up to mark

camsoft2000 的头像
camsoft20006 天前

Honestly, none. I had Claude research first: award-winning Mac and iOS utilities (Apple Design Award winners, CleanMyMac X, SoundSource 6, Linear) plus Apple's Passwords and Reminders, written up with real screenshots. The redesign grew out of that: one library modelled on Passwords, with its own colour for every destination. For the icon I pointed Claude at Matthew Skiles' work. The skeuomorphic skill cartridge, striped in those same colours, ties the whole theme together.

Bharath 的头像
Bharath6 天前

Amazing. Thanks for the details.

Yolo 的头像
Yolo6 天前

A tool to manage the skills my agent loads unnecessarily. Now I just need a tool to manage the tools.

Mateo Lafalce 的头像
Mateo Lafalce6 天前

cool, does toggling a skill off remove it from the current session's context, or do you need to start a new session for it to take effect?

camsoft2000 的头像
camsoft20006 天前

Mostly likely need to start a new session. It this is good practice anyway when changing directions.

Luis 的头像
Luis6 天前

Looks really cool! One thing I'd love: per-project skill sets, so two agents working in different repos at the same time can each have their own active group, instead of one global switch. Any plans for that?

Automater 的头像
Automater6 天前

Eager-loading every skill description is silent context tax — and conflicting skills fight for the same tokens. Selective activation beats a bigger skill folder: load what the task needs, park the rest. #ClaudeCode #DevTools

相关视频

Skills are the quickest way to 10x the quality and consistency of what you get from Claude Code. And you don't need to be a developer to use them. Anthropic just published how they use hundreds of skills internally every day. Most skill tutorials are made for developers — if you're in marketing, sales, content ops, or GTM, you probably watched those and moved on. But skills are just as important for non-developers. A skill is just a reusable prompt with clear instructions for a specific task. Instead of prompting Claude the same way over and over, you build it once and invoke it every time. I have a skill for writing on LinkedIn. A different one for YouTube outlines. Another for X. Each platform has different rules, different voice, different structure — so each one gets its own skill. If you're doing something repeatedly, it's time to make a skill. The biggest mistake most people make: building skills as a single .md file. A single file dumps everything into context whether Claude needs it or not. Wastes tokens. Gets worse results. Skills should be folders. Here's the structure that works: skill.md — the orchestrator. Tells Claude which files to read and when. It doesn't contain rules itself — it's the playbook. instructions/ — separate files for voice, structure, scope. Claude only loads the one it needs for the current step. examples/ — good AND bad. Good examples show what success looks like. Bad examples show patterns to avoid — AI writing tells, weak hooks, generic CTAs. Most people skip bad examples. Don't. eval/ — a checklist that scores every output before you see it. "Does it have a clear hook?" "Is it free of AI buzzwords?" Pass or fail on each item. templates/ — output formatting so you get consistent structure every time. The three types of skills that matter most for non-developers: 1. Business automation. Writing a newsletter. Checking reports and drafting follow-ups. Running programmatic ad campaigns. Any workflow you repeat — build a skill for it. 2. Content templates. Landing page copy, meta ads, email sequences, SEO briefs. Each one has specific requirements. Each one gets its own skill. 3. Thinking partners. This is the one people miss. Skills don't have to produce output. They can help you think — an advisory board that reviews your work from your ICP's perspective, a coach that pressure-tests your strategy, an ideation partner that researches competitors before suggesting your next move. If you already have skills as .md files, here's the exact prompt to restructure them in the Anthropic approved format: "I want to restructure my Claude Code skill file. Right now my skill is a single .md file and I want to break it into a folder system following Anthropic's best practices. Read my current skill file, then restructure it into a folder with: a skill.md orchestrator, an instructions/ folder with separate files for each concern (voice, structure, scope), an examples/ folder with good and bad examples, an eval/ folder with a quality checklist, and a templates/ folder for output formatting. Keep all my existing rules and intent — just reorganize them into the modular structure." Paste that into Claude Code pointed at the folder where your skill lives. It handles the rest. A few caveats: 1. Don't add too many skills. Every skill adds context Claude has to process. 50 skills loaded means everything slows down. Start with 3-5 covering your most repeated workflows. 2. Vet skills before downloading. If you grab a skill from the internet, read what's inside first. Skills can include shell commands and scripts. Check what you're running. 3. Share what works. Build a skill that performs well, put it in a shared GitHub repo. Your marketing org gets shared skills for copywriting, SEO, ad copy — new hires invoke the skill instead of learning every playbook from scratch. Onboarding time drops dramatically. 4. Keep your skills updated. When you see output you love, add it as a good example. When you see a pattern you hate, add it as a bad example. The skill gets sharper every time. I made a full video walking through all of this — including a live build of two skills from scratch (no terminal, no code), the exact prompt I use to restructure old skills, and 5 pro tips from Anthropic's internal playbook. Share this with your non-developer friends that want to do more with AI; or bookmark it to come back to at a later time.

JJ Englert

29,322 次观看 • 6 个月前