Feedback
What about a system that manages user API keys using a structured approach, similar to the Media Vault's folder layout? This system would store keys for platforms like OpenAI, MidoriAI, Groq, and Google. With sub folders like TTS, LRM, PhotoGen, VideoGen, MusicGen (yes Google has that) that are only made after a key is added. It would leverage the same encryption method as the Media Vault for secure key handling, with the keys stored in RAM (for a time, ie 15 mins between requests) to minimize disk usage.
When requesting a key we could have a key tester, that checks the key with the endpoint to make sure it is working. If the key is not working put it on cooldown.
Area of Feedback
Packaging / Distribution
Additional Details
Here's a possible use case:
Use Case: API Key Management for Rate Limit Optimization
As a Midori AI user, I have 15 API keys, each allowing only 2 requests per hour. To effectively utilize these keys and avoid rate limit issues, I propose implementing a key manager system. This system would:
- Dynamically load an API key based on usage patterns and the current state of each key.
- Seamlessly integrate with the agents system.
This approach would ensure optimal API key utilization and prevent disruptions due to rate limiting.
Feedback
What about a system that manages user API keys using a structured approach, similar to the Media Vault's folder layout? This system would store keys for platforms like OpenAI, MidoriAI, Groq, and Google. With sub folders like TTS, LRM, PhotoGen, VideoGen, MusicGen (yes Google has that) that are only made after a key is added. It would leverage the same encryption method as the Media Vault for secure key handling, with the keys stored in RAM (for a time, ie 15 mins between requests) to minimize disk usage.
When requesting a key we could have a key tester, that checks the key with the endpoint to make sure it is working. If the key is not working put it on cooldown.
Area of Feedback
Packaging / Distribution
Additional Details
Here's a possible use case:
Use Case: API Key Management for Rate Limit Optimization
As a Midori AI user, I have 15 API keys, each allowing only 2 requests per hour. To effectively utilize these keys and avoid rate limit issues, I propose implementing a key manager system. This system would:
This approach would ensure optimal API key utilization and prevent disruptions due to rate limiting.