LIVE
Anthropic says Claude helped trace the origin of NIST's mysterious ECDSA seeds16/10/26 · Anthropic|Reflection AI unveils Beam, a 501-billion-parameter open model16/10/26 · Reflection AI|Meta releases Muse Spark Open, a mixture-of-experts model built for local Mac Studio inference15/10/26 · Meta|Meta releases Muse Spark Open, an open-weight multimodal model15/10/26 · Meta|Google DeepMind outlines safety measures behind Gemini 4 Argon13/10/26 · Google DeepMind|Tev-Gym: A Training Recipe to Fix Sparse Rewards in Long-Horizon Agents13/10/26|Google DeepMind references Gemini 4, code-named Argon13/10/26 · Google DeepMind|Google DeepMind unveils Gemini 4 Argon, a robotics-focused variant13/10/26 · Google DeepMind|Character.AI files for Chapter 11 bankruptcy, sells assets to Disney13/10/26 · Character.AI|Study Shows AI Models Can Provably Conceal Arbitrary Capabilities13/10/26|NVIDIA RTX Spark laptops bring local AI to premium Windows PCs17/10/26 · NVIDIA|OpenAI discloses four more incidents involving sensitive data17/10/26 · OpenAI|Anthropic says Claude helped trace the origin of NIST's mysterious ECDSA seeds16/10/26 · Anthropic|Reflection AI unveils Beam, a 501-billion-parameter open model16/10/26 · Reflection AI|Meta releases Muse Spark Open, a mixture-of-experts model built for local Mac Studio inference15/10/26 · Meta|Meta releases Muse Spark Open, an open-weight multimodal model15/10/26 · Meta|Google DeepMind outlines safety measures behind Gemini 4 Argon13/10/26 · Google DeepMind|Tev-Gym: A Training Recipe to Fix Sparse Rewards in Long-Horizon Agents13/10/26|Google DeepMind references Gemini 4, code-named Argon13/10/26 · Google DeepMind|Google DeepMind unveils Gemini 4 Argon, a robotics-focused variant13/10/26 · Google DeepMind|Character.AI files for Chapter 11 bankruptcy, sells assets to Disney13/10/26 · Character.AI|Study Shows AI Models Can Provably Conceal Arbitrary Capabilities13/10/26|NVIDIA RTX Spark laptops bring local AI to premium Windows PCs17/10/26 · NVIDIA|OpenAI discloses four more incidents involving sensitive data17/10/26 · OpenAI|
+ nAIvigateA SHEET FOR PEOPLE WITHOUT 2 HOURS
+ nAIvigate · fiches / 40-outils-claude-code
▣ fiche-002.md ×
□ index.md
☰ ContentsOfficial foundationHarnessMemory & contextToolsSecurityCost

The 40 things to install in Claude.

Claude Code ships with almost none of this. Here are the 39 repos that matter, across six families, including security, which every other list forgets. Every star count is pulled live from the GitHub API each time this page is updated, not copied from a screenshot.

Official foundation

the Anthropic & protocol foundations5 rows
skills ↗The official skills library. The starting point. anthropics/skills★ 170k
claude-code ↗The product's repo: changelog, issues, discussions. anthropics/claude-code★ 141.8k
claude-cookbooks ↗Official recipes and notebooks, from the API to agents. anthropics/claude-cookbooks★ 51.7k
servers ↗The reference MCP servers, maintained by the protocol. modelcontextprotocol/servers★ 89.6k
github-mcp-server ↗The official GitHub MCP: reads and writes to your repos. github/github-mcp-server★ 32.3k

Harness

what it does before typing7 rows
superpowers ↗Plans, TDD, debugs before writing. Accepted into the official marketplace. obra/superpowers★ 273.2k
ECC ↗The firehose aggregator: configs, hooks, commands. affaan-m/ECC★ 240.7k
awesome-claude-code ↗The canonical curated list of the ecosystem. hesreallyhim/awesome-claude-code★ 52.5k
claude-task-master ↗Breaks a specification down into executable tasks. eyaltoledano/claude-task-master★ 28k
ruflo ↗Multi-agent swarm orchestration. ruvnet/ruflo★ 68.1k
agents ↗Marketplace of ready-to-use sub-agents. wshobson/agents★ 38.9k
claude-code-templates ↗Configuration and monitoring CLI, with hundreds of templates. davila7/claude-code-templates★ 30.3k

Memory & context

what survives beyond the session6 rows
repomix ↗Packs an entire repo into a single file for context. yamadashy/repomix★ 27.9k
claude-mem ↗Persistent context from one session to the next. thedotmack/claude-mem★ 91k
mem0 ↗Universal memory layer for agents. mem0ai/mem0★ 63.5k
context7 ↗Up-to-date docs, exact version, injected into context. upstash/context7★ 60.9k
awesome-claude-skills ↗Curated skills: memory, workflow, productivity. travisvn/awesome-claude-skills★ 14.7k
awesome-claude-plugins ↗15,000+ plugins indexed with adoption metrics. quemsah/awesome-claude-plugins★ 1.2k

Tools

beyond the terminal11 rows
firecrawl ↗Scrapes and searches any website, cleanly. firecrawl/firecrawl★ 168.6k
playwright-mcp ↗Tests in a real browser. microsoft/playwright-mcp★ 36.2k
chrome-devtools-mcp ↗Debugs in Chrome with real DevTools. ChromeDevTools/chrome-devtools-mcp★ 49.3k
browser-use ↗Makes websites controllable by your agents. browser-use/browser-use★ 109.5k
vibe-kanban ↗A board to track what your agents are doing. BloopAI/vibe-kanban★ 27.8k
claude-code-router ↗Routes each task to the right model. musistudio/claude-code-router★ 36.7k
cc-switch ↗Switches Claude ⇄ Codex and more. farion1231/cc-switch★ 127.8k
opcode ↗Desktop GUI to drive Claude Code. winfunc/opcode★ 22.4k
mcp ↗Your Supabase database driven via MCP. supabase/mcp★ 2.9k
mcp-server-cloudflare ↗Workers, DNS, KV in natural language. cloudflare/mcp-server-cloudflare★ 4.1k
awesome-mcp-servers ↗All the MCP servers, listed. punkpeye/awesome-mcp-servers★ 92.5k

Security

the section everyone else forgets6 rows
semgrep ↗Static scan before the agent commits. semgrep/semgrep★ 16.3k
gitleaks ↗Hunts for secrets your agents might push. gitleaks/gitleaks★ 28.8k
trivy ↗Vulnerabilities in dependencies and containers of generated code. aquasecurity/trivy★ 37.5k
tdd-guard ↗Enforces TDD: no code without a failing test first. nizos/tdd-guard★ 2.3k
garak ↗LLM vulnerability scanner: jailbreaks, leaks, injections. NVIDIA/garak★ 8.8k
system-prompts-and-model… ↗The real system prompts of AI tools, collected. x1xhlol/system-prompts-and-models-of-ai-tools★ 142.9k

Cost

what it costs to run4 rows
ccusage ↗Claude Code spend, analysed from your logs. ccusage/ccusage★ 18k
Claude-Code-Usage-Monitor ↗Real-time tracking of tokens and burn rate. Maciek-roboblog/Claude-Code-Usage-Monitor★ 8.6k
ccstatusline ↗Cost and context displayed in the status bar. sirmalloc/ccstatusline★ 12.4k
claude-quickstarts ↗Official starter apps so you don't pay for exploration. anthropics/claude-quickstarts★ 17.5k

Go deeper

fundamentals, in Learn2 links
These tools are already running in our stack. Automation Sprint: we industrialise this stack in your context, with ROI measured at 90 days.
TALK AUTOMATION →
⎇ mainfiche #002 · outilssources: API GitHub (live) · anthropics/skills · awesome-mcp-servers
Follow nAIvigate | naivigate.com/en/fiches