← ClaudeAtlas

using-context7listed

Use when working with any library/framework/SDK to fetch current docs. Beats training-data memory which is stale.
liujiarui0918/claude-code-codex-strongest · ★ 0 · AI & Automation · score 62
Install: claude install-skill liujiarui0918/claude-code-codex-strongest
# Using Context7 ## What this is Context7 is an MCP server that fetches **current** documentation for libraries, frameworks, and SDKs on demand. Your training data is months-to-years stale; Context7 is the corrective. ## When to use Use whenever the user asks about: - A library's API, configuration, or setup (`React`, `Next.js`, `Django`, `Prisma`, `Tailwind`, `Express`, `Spring`, `FastAPI`, `LangChain`, ...) - Version-specific behavior or migration (`Next.js 14 → 15`, `React 18 → 19`) - "How do I do X with Y" where Y is any named library/framework - Best practices or recommended patterns for a specific tool - Configuration files (`tailwind.config.js`, `next.config.ts`, `prisma.schema`) **Even if you "feel like you know" the answer** — query Context7 first. Library APIs churn fast: server actions, app router, RSC, Tailwind v4, Prisma 6 migrations all shipped after most training cutoffs. Confidence-from-memory is the failure mode. ## When NOT to use - Refactoring existing code (the code itself is the spec) - Writing PowerShell/Bash scripts (general shell knowledge) - Debugging business logic (no library API involved) - Code review for style/correctness - Generic programming concepts (sorting algorithms, OOP patterns, recursion) - Internal company code (Context7 only knows public libraries) ## Workflow Two-call pattern, always in this order: ### Step 1: Resolve library ID ``` mcp__context7__resolve-library-id libraryName: "Next.js" # official name with p