wille/haprovider — reverse-engineered prompt

Build me a small production ready service that sits in front of my Ethereum and Solana RPC providers and gives my apps one stable endpoint to use. It should handle normal JSON RPC over HTTP and websockets, send requests to whichever upstream node is healthy, automatically fail over when a provider times out, errors, or rate limits me, and start using providers again when they recover.

I want a simple yml config file, plus env or command line overrides, so I can define named endpoints like Ethereum mainnet with a chain id and a few providers, and Solana with a couple of backups. Please include health checks, request and response validation, connection reuse, useful logs, and Prometheus metrics so I can see traffic, failures, latency, open connections, and provider health. Make it easy to run locally or in Docker with sensible defaults around port 8080. Keep the scope focused on reliability and observability, not caching or advanced transaction tricks. Look up current docs online if you need to.