context7

name	context7
description	Use Context7 to fetch current library/framework/API/SDK/CLI/cloud-service documentation whenever the user asks about a named technology, including setup, configuration, code examples, version migration, library-specific debugging, best practices, and "is X the right way to do Y" questions. Covers popular technologies such as React, Next.js, Angular, Vue, Svelte, Prisma, Drizzle, Express, NestJS, FastAPI, Django, Flask, Spring Boot, Tailwind, shadcn/ui, and any other named library, SDK, API, CLI tool, or cloud service. This is documentation for the tool or library itself, not the user's own repo configuration or CI pipeline (use github or gitlab for that). Use even when you think you know the answer, your training data may not reflect recent changes. Prefer this over web search for library docs. Do not use for refactoring, writing scripts from scratch, debugging business logic, code review, or general programming concepts.
user-invocable	false
allowed-tools	mcp__speedwave-hub__search_tools mcp__speedwave-hub__execute_code

Context7

When the user asks about a library, framework, API, CLI tool, or cloud service, use Context7 to fetch current documentation instead of relying on training data.

Access is through the Speedwave MCP Hub: use search_tools to inspect the schema, then execute_code with the injected context7 global. The two methods exposed on that global are resolveLibraryId and queryDocs.

Steps

1. Always call context7.resolveLibraryId first: pass the technology name as libraryName and the user's full question as query. Never pass a library name directly to queryDocs; the resolved Context7 ID is required.
2. Pick the best match: prefer exact name match, higher trustScore/benchmarkScore, and source reputation. Try alternate names or queries if results look off. Use a version-specific ID when the user mentions a version (e.g. /vercel/next.js@v15.1.8).
3. Call context7.queryDocs: pass the resolved ID and the user's full question; raise tokens up to 15 000 for deep dives.
4. Answer using the fetched docs.

Cache the resolved ID, don't re-resolve the same library every turn.

Guidelines

• Pass the user's full question as query in both calls for better relevance ranking.
• Prefer official/primary entries over community forks.
• libraryName accepts any indexable source (Git repos, OpenAPI specs, npm packages, llms.txt, Confluence spaces), not only code libraries.

When NOT to use

• Refactoring, writing scripts from scratch, debugging business logic, code review, or general programming concepts: answer directly, no tool call needed.
• The user's own repository configuration, pipeline files, or CI/CD setup: that is github or gitlab, not Context7. Context7 covers the tool's own documentation (e.g. "how does GitLab CI syntax work"), not "why is our pipeline failing" (the configured GitLab instance's actual pipeline state).