mirror of
https://github.com/diegosouzapw/OmniRoute.git
synced 2026-08-03 13:52:09 +03:00
7 lines
2.5 KiB
JSON
7 lines
2.5 KiB
JSON
{"prompt":"Hi there, I would like you to provide a detailed review of the authentication middleware because I think that it is being used to validate API keys, and also it should protect the dashboard routes. Additionally, please explain in detail how the authentication middleware should report unauthorized requests without leaking secrets. Thanks in advance.","keyPhrases":["authentication middleware","API keys","dashboard routes","unauthorized requests","secrets"]}
|
|
{"prompt":"Hello, I was wondering if you could create a thorough migration plan for the SQLite migration that adds a cavemanConfig key. The goal is to keep existing compression settings, and also preserve user overrides, in order to avoid regression during release. Please include the SQLite migration number and cache invalidation behavior.","keyPhrases":["SQLite migration","cavemanConfig","compression settings","user overrides","cache invalidation"]}
|
|
{"prompt":"Could you show me how the rate limit reset should work when the provider returns a retry-after header? Basically the rate limit reset must update the account cooldown, and also it should not poison other provider accounts. I really appreciate a concise explanation with the exact rate limit reset flow.","keyPhrases":["rate limit reset","retry-after header","account cooldown","provider accounts"]}
|
|
{"prompt":"Here is my code:\n```ts\nexport function routeRequest(model: string) {\n return model.startsWith(\"combo/\") ? \"combo\" : \"single\";\n}\n```\nPlease explain in detail how this routing helper should keep the combo routing behavior while also allowing single model fallback. The key requirement is preserving combo routing behavior.","keyPhrases":["routeRequest","combo routing behavior","single model fallback"]}
|
|
{"prompt":"As we discussed earlier, the cache invalidation path is very important because the settings page updates compression defaults, and also the request pipeline reads those defaults before translating provider payloads. I need you to explain why cache invalidation must happen after writing key_value rows.","keyPhrases":["cache invalidation","settings page","compression defaults","request pipeline","key_value rows"]}
|
|
{"prompt":"I am trying to debug a retry queue that was implemented for background jobs. Furthermore, the retry queue should keep failed jobs visible, and also it should not retry completed jobs. Please provide a detailed summary of how the retry queue should record failure state and job visibility.","keyPhrases":["retry queue","background jobs","failed jobs","completed jobs","failure state","job visibility"]}
|