feat: add Requesty as an OpenAI-compatible provider

[Thibaultjaigu]

This adds a dedicated requesty provider, mirroring the existing OpenRouter provider as closely as possible.

Changes:

• backend_api_python/app/services/llm.py: add REQUESTY to the LLMProvider enum, a PROVIDER_CONFIGS entry (base URL https://router.requesty.ai/v1), the get_api_key key map, and the auto-detect priority list.
• backend_api_python/app/config/api_keys.py: add a REQUESTY_API_KEY property mirroring OPENROUTER_API_KEY (env var first, then addon config).
• backend_api_python/app/utils/config_loader.py: add the REQUESTY_API_KEY to requesty.api_key env mapping.

Requesty is an OpenAI-compatible LLM gateway and uses the same provider/model naming as OpenRouter, so it reuses the same OpenAI-compatible client path (no new request code).

Verification: I tested the endpoint live before opening this. GET https://router.requesty.ai/v1/models returned 200 (around 578 models), and a chat completion with openai/gpt-4o-mini against https://router.requesty.ai/v1/chat/completions returned 200. I confirmed the enum, config, key map, and priority list all resolve requesty consistently.

Docs: https://requesty.ai , https://app.requesty.ai/api-keys

I work at Requesty. This mirrors the existing OpenRouter provider as closely as possible. Happy to adjust or close it if it is not a fit.