Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Claude Code feels completely different once you install this. Anthropic quietly released an official plugin called claude-code-setup and it basically turns Claude Code from “pretty good” into an actual AI dev environment. It scans your project and recommends: → hooks → skills → MCP servers → subagents → automations...

75,447 Aufrufe • vor 5 Monaten •via X (Twitter)

0 Kommentare

Keine Kommentare verfügbar

Kommentare vom Original-Post werden hier angezeigt

Ähnliche Videos

NOBODY IS F*CKING TALKING ABOUT THIS Anthropic just dropped a treasure chest of 21 Claude Code plugins. $43,559 worth of tools, FREE on GitHub paid marketplaces charge $19 for a skill and $39 for an agent. these plugins ship 1,100+ skills and 580+ agents for $0 1.2M stars combined, picked from Anthropic's own store of 300+ plugins and the best of GitHub most people still run Claude Code blind: no usage tracking, no reviews, no design taste, burning tokens reading whole files 21 tools you should try first: 01 ccusage → see exactly where your Claude Code money goes: per day, per session, per model and per 5-hour window 02 claude-context → semantic code search for big repos. ~40% fewer tokens, same retrieval quality 03 Serena → finds and edits code by symbol, so Claude stops reading whole files to change one function 04 claude-code-router → one local endpoint that routes Claude Code to DeepSeek, Kimi, Gemini or OpenRouter 05 session-report (Anthropic) → an HTML report of your tokens, caching and most expensive prompts 06 claude-code-templates → 420+ agents, 890+ skills and 280+ commands in one npx install 07 awesome-claude-code-subagents → 161+ ready subagents across 10 categories, install by category 08 scientific-skills → 177 research skills plus ~100 scientific databases, from bioinformatics to materials 09 superpowers → brainstorm, plan, TDD and systematic debugging, so Claude works like a senior and not an intern 10 Context7 → pulls up-to-date, version-specific docs into Claude Code, so it stops writing outdated or made-up API code 11 Firecrawl → your agent searches the web, scrapes any page into clean data and works with live sites 12 browser-use → gives Claude Code a real browser that clicks, fills forms and finishes tasks on live sites 13 claude-code-setup (Anthropic) → scans your repo and tells you which hooks, skills, MCP servers and subagents to add 14 code-review (Anthropic) → several agents review your PR and score every finding by confidence 15 security-guidance (Anthropic) → warns on risky edits and checks every commit for security holes 16 frontend-design (Anthropic) → production-grade UI with real design quality, not the usual AI look 17 hookify (Anthropic) → turns "never do that again" into a real hook, straight from the chat 18 Spec Kit (GitHub) → spec-driven development: write the spec first, Claude Code builds from it step by step 19 MCP servers → the official reference servers: filesystem, git, memory and fetch for your agent 20 Task Master → breaks your project into tasks and runs them through Claude Code, no API key needed 21 shopify-ai-toolkit (Shopify) → 18 skills for docs, API schemas, code validation and store management start with 3, not 21: one to track usage, one to cut tokens, one for your actual stack run one real task before and after. keep only what makes it faster or cheaper every plugin you install eats context, so stack them carefully save this before you build a tool that's already sitting here then read how to stop the 9 leaks draining your Claude limit in the article below ↓

chuplung

16,673 Aufrufe • vor 1 Tag