notBlubbll/free-buff-lol ? reverse-engineered prompt
Reverse engineered prompt
Build me a Node.js proxy app for Freebuff that makes it feel like a normal OpenAI and Anthropic API server.
I want to point other tools at it and use /v1/chat/completions, /v1/models, /v1/messages, and /v1/messages/count_tokens, with streaming working too. It should handle multiple auth tokens, rotate between them, manage sessions automatically, and retry cleanly when rate limits happen. If a model is locked or not allowed for the current session, it should recover on its own and pick a working option.
Also include a simple dashboard page where I can see the available models, toggle them on or off, and handle the login or OAuth flow. It should auto configure itself on startup if possible, and keep the model list in sync with upstream. If you need current details from the docs or upstream sources, look them up online and make it work end to end.
Are you gonna build this?
make sure you review the code using coderabbit