Claude Code /skill-doctor: 64 Unused Skills Cost Under 1% of My Context

Claude Code /skill-doctor shows what each of your Claude Code skills costs in context and when you last used it. On my setup it flagged 64 never-used skills, and switching them all off saves under 1% of my context. The weight sits in skills I did use, and the report doesn't flag those. ---- ❤️ SUPPORT THE CHANNEL Patreon: https://www.patreon.com/cw/DIYSmartCode Higgsfield AI — Create AI Videos & Visuals Get 10% OFF here: https://higgsfield.ai?fpr=thomas-dccf7e 🚀 DYNAMOUS AI COMMUNITY Want to learn agentic coding with live daily events and workshops? Check out Dynamous AI: https://dynamous.ai/?code=646a60 Get 10% off here 👉 https://shorturl.smartcode.diy/dynamous_ai_10_percent_discount ⚡ HOSTINGER — RELIABLE HOSTING FOR YOUR PROJECTS (10% OFF) Whether you're shipping a portfolio, a side project, n8n flows, or AI agents — I use Hostinger for fast, affordable VPS + web hosting. Get 10% off here 👉 https://hostinger.com/DIYSMARTCODE n8n: https://n8n.partnerlinks.io/diysmartcode (Affiliate links — costs you nothing, supports the channel.) ---- What you will see in this 2-minute breakdown: • Where the report shows up: the Stats tab in /plugin in an interactive session, plain text with claude -p "/skill-doctor" (real run, Claude Code 2.1.310, on the repo I build these videos with) • What each entry shows: the context cost of the skill's one-line listing on every turn, the uses, and when you last used it; a dash means the skill isn't in the listing and costs nothing • The run: 257 skills loaded, over half show a dash, 64 flagged as never invoked, each one at 20 tokens or less, about 1,300 tokens together against a session that starts at 155k • Why they're so cheap: the skill listing has a budget of 1% of the context window, and when it runs over, Claude Code drops descriptions from your least used skills first • The heavy entries: three skills that ran exactly once cost about 300 tokens each on every turn, and my most used skill (213 uses) hasn't run in 88 days; skill doctor flags neither • Skills synced from claude.ai: a deleted local copy is re-downloaded on the next sync, so turn those off on claude.ai • When the command isn't available: sessions that skip feature-flag fetching, for example with DO_NOT_TRACK or DISABLE_TELEMETRY set, or on Amazon Bedrock • The report also lists plugins you haven't used recently, and leaves out bundled and enterprise skills • Version: the docs say /skill-doctor needs Claude Code v2.1.252 or later; the changelog lists it under 2.1.261 Claude Code skills docs (the /skill-doctor report, listing budget): https://code.claude.com/docs/en/skills Claude Code commands reference (/skill-doctor): https://code.claude.com/docs/en/commands Claude Code environment variables (feature-flag fetching): https://code.claude.com/docs/en/env-vars Claude Code error reference (/skill-doctor in non-interactive mode): https://code.claude.com/docs/en/errors Claude Code changelog (2.1.261, /skill-doctor added): https://github.com/anthropics/claude-code/blob/main/CHANGELOG.md A skill you haven't used in a month: switch it off, or keep it just in case? #ClaudeCode #Claude #ClaudeAI #Anthropic #ClaudeCodeSkills #ClaudeSkills #SkillDoctor #AgentSkills #ContextWindow #ClaudeCodeTutorial #ClaudeAITutorial #TokenUsage #AICoding #AIAgents #AgenticCoding #DeveloperTools #CodingTools #VibeCoding #ClaudeBasics #Shorts