Загрузка видео...

Не удалось загрузить видео

На главную

Claude Code just shipped a new /skill-doctor command It shows which skills are loaded in your session, which ones haven’t been used, and how many context tokens each is costing you Now you know exactly what to remove

70,188 просмотров • 1 месяц назад •via X (Twitter)

Комментарии: 26

Фото профиля suqitah
suqitah1 месяц назад

the unused skill list is the real feature, most people's setup is just cruft nobody prunes

Фото профиля Yuv Bindal
Yuv Bindal1 месяц назад

tracking token cost per skill is clever, let's connect soon

Фото профиля Luis
Luis1 месяц назад

After using Astra yesterday, I'm starting to think figuring out which instructions to remove matters as much as knowing what to add. I'm doing a full cleanup of my skills and agent instructions now. Astra in Codex feels like a new platform.

Фото профиля Gabriele Spata
Gabriele Spata1 месяц назад

cost visibility is good, it just answers what's expensive, not what's working

Фото профиля 9h0st
9h0st1 месяц назад

skill-doctor 最有用的输出不只是“删哪些技能”,而是按触发频率/上下文成本/实际收益排序;高 token 但低命中率的规则,应该先懒加载或拆到子技能。

Фото профиля Ehsan Azish
Ehsan Azish1 месяц назад

a command whose whole job is telling you what to delete, that's a first for a cli tool

Фото профиля Tri
Tri1 месяц назад

cc @wjayesh

Фото профиля elian
elian1 месяц назад

pr #867 for a slash command, this repo moves faster than most seed stage roadmaps

Фото профиля Hussain Hashim | Building SundayBack
Hussain Hashim | Building SundayBack1 месяц назад

@dani_avila7 love it! it’s like decluttering for your AI setup. this would've saved me so much time debugging last month!

Фото профиля Ned Palmer
Ned Palmer1 месяц назад

Now if only I could actually delete shared skills, that would be great. Also better organization of org level skills so you can limit them to LOBs not the entire firm.

Фото профиля Marwan Atef
Marwan Atef1 месяц назад

skill-doctor is the audit every agent stack was missing. load list, unused skills, token burn per skill. finally a reason to prune the skill graveyard instead of pretending it is free context.

Фото профиля Catalin
Catalin1 месяц назад

Great thanks

Фото профиля Umang
Umang1 месяц назад

“Last used: never” as the default row is a pretty good roast of most Claude Code setups.

Фото профиля weigi
weigi1 месяц назад

this is such a good diagnostic — seeing unused skills and their context cost makes cleanup way less guessy

Фото профиля Vishnu Aniyan
Vishnu Aniyan1 месяц назад

Is this the same as /doctor ?

Фото профиля AI Mastery Guide
AI Mastery Guide1 месяц назад

skill doctor is exactly what i needed

Фото профиля Saikat | GoShipped
Saikat | GoShipped1 месяц назад

This is a good feature to have. Spending unnecessary tokens is like a death warrant.

Фото профиля Silk Song
Silk Song1 месяц назад

Why is the command reporting usage reports are not available on this connection

Фото профиля Mircea Sandu
Mircea Sandu1 месяц назад

nice! maybe this will finally convince me to try removing all my skills and adding them back as needed

Фото профиля 👽 John Mac
👽 John Mac1 месяц назад

Quite lovely Sir! Been waiting for this. Thanks for being a news channel worth paying attention to.

Фото профиля Chaminda
Chaminda1 месяц назад

unused this session and unused are different things. the skill you only load during an incident will look like dead weight.

Фото профиля Bessi with black hoodie
Bessi with black hoodie1 месяц назад

I've left dead skills loaded for weeks. /skill-doctor is the cleanup I needed.

Фото профиля Cikocisco
Cikocisco1 месяц назад

Handy. But does it flag skills silently bloating context without pulling weight? Or just list 'em?

Фото профиля MARUF
MARUF1 месяц назад

pruning unused context tokens saves money and avoids hallucinations

Фото профиля John Rood
John Rood1 месяц назад

unused in one session doesn't mean removable. rare guardrails and dead weight look identical at the token-count level. the doctor needs to show which skill prevented which failure.

Фото профиля Monaal
Monaal1 месяц назад

token count is only the first part of what a skill costs. Cached and uncached input bill at different rates, so a large skill sitting in a stable prefix is far cheaper than the raw number suggests. And a rarely-used one may be preventing a failure that only shows up downstream.

Похожие видео

THIS MIGHT BE THE #1 OPEN-SOURCE REPO FOR CLAUDE CODE RIGHT NOW. IT GIVES CLAUDE A MEMORY AND SLASHES YOUR TOKEN COST ON EVERY QUESTION The repo is safishamsi/graphify, a free open-source skill that turns any codebase into a knowledge graph Claude Code can read instantly. Instead of grepping through your files every session, Claude gets a map of how everything connects The problem it fixes: Every time you ask Claude Code about a big repo, it does the same thing, greps through dozens of files like a brute-force Ctrl+F, blows through your context window, and sometimes still misses the answer hiding in a file nobody searched. Claude Code has no memory of how your project is structured. Every session starts from zero What it does: It maps your entire codebase into a knowledge graph, capturing not just which files exist, but which functions depend on which, which modules are central, and which files cluster around the same concern. Claude queries the map instead of scanning files How it works, three passes: 1. Code structure, free and local. Tree-sitter parses your files and pulls out classes, functions, imports and call graphs. No LLM, no tokens, just your actual code mapped deterministically 2. Audio and video, if you have them. Transcribed locally and folded into the graph 3. Docs, papers, images. Here an LLM does semantic analysis, figuring out what each document means and where it fits. Only the meaning gets sent up, never your raw source It saves you money: Normally a question about a big repo makes Claude spawn explore agents that scan file after file, eating your context window and your token budget before you get an answer. With the graph already built, Claude queries the map instead of re-reading the codebase every time. Same answer, a fraction of the tokens. The graph only gets built once, then a hook rebuilds it after each commit for free, so you never pay that scanning cost again. The bigger the repo, the bigger the gap The best parts: it's a skill, so once installed Claude knows when to use it without you memorizing commands. It works on non-code folders too, point it at docs or notes and it can spin up an Obsidian vault How to add it to your Claude: 1. Install Claude Code if you haven't: npm install -g Paul Jankura-ai/claude-code 2. Add the skill: claude skill add safishamsi/graphify 3. Open your project folder and run /graphify . to build the graph 4. Optional, make it automatic: graphify hook install so the graph rebuilds after every commit That's it. Ask Claude about your repo and it reads the map instead of burning tokens on a file hunt Bookmark this

Yarchi

56,568 просмотров • 4 месяцев назад

this is pure f*cking treasure Anthropic devs just posted 20 GitHub repos that turn your Claude Code into a f*cking engineering team: Exposed for Opus 5.5 1. rtk: squeezes command output before it hits your context. the authors claim 60-90% fewer tokens 2. claude-mem: memory across sessions. records what the agent did and brings back what matters 3. ccusage: shows where your tokens and money went. one command, npx ccusage 4. Serena: search and edit code by meaning, not by text 5. Context7: up-to-date library docs, so Claude stops writing for last year's API 6. claude-context: your whole codebase searchable through MCP 7. Repomix: packs an entire repo into one AI-friendly file skills and method 8. superpowers: a skills framework and a full development methodology. the one almost everyone installs 9. skills: the official Agent Skills repo from Anthropic 10. spec-kit: spec-driven development, from GitHub itself 11. agents: a plugin marketplace for Claude Code and six other harnesses 12. claude-code-templates: ready configs and monitoring from one CLI 13. claude-code-hooks-mastery: hooks explained with working examples 14. awesome-claude-code: the hand-picked map of everything else running more than one agent 15. container-use: a separate container for every agent, so they stop breaking each other 16. claude-squad: several terminal agents in one window 17. Happy: control your agents from your phone, iOS and Android 18. claude-code-router: route requests across different models 19. ccstatusline: a status line that shows what your session is doing 20. tweakcc: edit Claude Code's own system prompts, toolsets and themes the stack in one line: memory keeps the context, skills set the method, containers let you run five agents instead of one. check each repo before you install. save this.

kiosa

70,094 просмотров • 24 дней назад