Instant pub.dev documentation API optimized for LLM agents. Lightning-fast responses with three-tier caching, semantic search, and structured data ready for AI coding tools.
The challenge of accessing Flutter documentation programmatically, and how PubContext is solving it.
pub.dev doesn't offer a REST API to fetch full package documentation including classes, methods, properties, and descriptions.
Both the documentation website and API enforce strict rate limits, making direct scraping impractical for on-demand fetching.
Extracting structured data requires navigating multiple pages per package, parsing complex DOM structures, and handling errors.
HTML documentation includes navigation, ads, and markup noise that wastes tokens and confuses AI models.
Memory cache for hot data, disk storage for persistence, and background freshness checks to keep docs current without blocking.
Go-powered with goroutines and worker pools, serving thousands of requests simultaneously without blocking.
Exponential backoff retry logic and request delays ensure reliable scraping without overwhelming pub.dev.
Cached responses return instantly, with async jobs for large packages and pre-warming of popular packages on startup.
Clean markdown or structured JSON with zero HTML noise, token-efficient chunking, and depth control.
RESTful endpoints with semantic search, package discovery, and helpful cache headers.
RESTful API with consistent response formats and helpful cache headers.
GitHub Copilot, Cursor, and other AI tools can use PubContext to get accurate Flutter package documentation for code generation.
Build agents that can research Flutter packages, understand APIs, and generate integration code with real documentation context.
IDE plugins, CLI tools, and documentation browsers can leverage PubContext for fast, structured package information.