Add Eden AI as a first-class LLM provider (and embedder)

[MVS-source]

Summary

Add Eden AI as a first-class LLM provider (and embedder) in repowise.

Eden AI is an EU-headquartered gateway that exposes 700+
models (Mistral, GPT, Claude, Gemini, Cohere, DeepSeek, Llama, …) behind a single
OpenAI-compatible endpoint (https://api.edenai.run/v3), with an EU endpoint
(https://api.eu.edenai.run/v3) for data residency / GDPR-sensitive workloads.

Why it fits repowise

• It's OpenAI-compatible, so it slots into the existing provider pattern the same
  way openrouter / deepseek do — no new dependency (reuses the openai SDK
  with a custom base_url).
• One key unlocks many vendors, which is convenient for the BYOK model, and the EU
  endpoint is useful for teams with data-residency requirements.

Proposed scope

• EdenAIProvider (llm/edenai.py) — generate() + streaming chat + /models
  discovery; default mistral/mistral-small-latest; EDENAI_API_KEY /
  EDENAI_BASE_URL; reasoning mapped to OpenAI reasoning_effort for OpenAI
  reasoning models routed via Eden.
• EdenAIEmbedder (embedding/edenai.py) — openai/text-embedding-3-small default.
• Registry + rate-limiter + server catalog + CLI selection/validation + web UI + tests + docs.

I've implemented this following .github/CONTRIBUTING.md and verified it end-to-end
against the live API. Happy to open the PR — filing this first per the issue-first
convention. Would you be open to it?