feat: add Requesty provider

[Thibaultjaigu]

Summary

Adds Requesty as an OpenAI-compatible LLM provider, implemented as a new RequestyProvider(BaseProvider) that mirrors the existing OpenRouterProvider as closely as possible.

Requesty is an OpenAI-compatible LLM gateway:

• Base URL: https://router.requesty.ai/v1
• Auth: Authorization: Bearer $REQUESTY_API_KEY
• Model naming: provider/model (same convention as OpenRouter, e.g. openai/gpt-4o-mini)

Implementation

ai/providers/requesty_provider.py copies the OpenRouter provider's structure:

• Reads ai_config.get("requesty", {})
• API key from config or REQUESTY_API_KEY env var
• Backend is ChatOpenAI(openai_api_base="https://router.requesty.ai/v1", ...)
• Default model openai/gpt-4o-mini
• generate_with_usage sets the response dict "provider" field to "requesty"
• All other logic (context formatting, sync/async generate, availability check) is identical to OpenRouter.

Wiring sites updated

• ai/providers/__init__.py — registered "requesty" in the in-tree PROVIDERS registry
• config/guardian.yaml — added a requesty: config block, updated the active-provider comment, and added a pricing entry
• cli/commands/scan.py, workflow.py, recon.py — added requesty to each valid_providers list and the --provider help text
• cli/commands/models.py — added a "REQUESTY MODELS" table section and the env-var hint
• tests/test_provider_plugins.py — added requesty to the in-tree provider assertion tuple
• README.md — added Requesty to provider lists/counts, config snippet, env-var examples, architecture/structure diagrams, and an API-key link

Verification

• Ran a live chat completion against https://router.requesty.ai/v1/chat/completions with model openai/gpt-4o-mini — received a successful 200 completion.
• Ran python -m pytest tests/test_provider_plugins.py — all 7 tests pass.

Links

• https://requesty.ai
• https://docs.requesty.ai

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