[Provider] Add Modelsell as an OpenAI-compatible LLM provider

Enhancement

Add Modelsell as a named OpenAI-compatible LLM provider in Continue.

Modelsell exposes the standard endpoints Continue needs for an initial integration:

• API base: https://modelsell.com/v1
• Chat Completions: POST /v1/chat/completions
• Model discovery: GET /v1/models
• Authentication: bearer API key
• API keys: https://modelsell.com/console/token
• Documentation: https://modelsell.com/docs/api-reference

Motivation

Modelsell can be configured through Continue's existing OpenAI-compatible path, but users must manually repeat the API base, credential setup, and model identifiers. A named provider would make the integration discoverable and give Continue a stable place for provider defaults without introducing a separate transport implementation.

Proposed implementation

1. Add a modelsell provider that reuses or extends the existing OpenAI Chat Completions implementation.
2. Default the API base to https://modelsell.com/v1 and use MODELSELL_API_KEY/the normal Continue secret mechanism.
3. Discover model IDs from GET /v1/models rather than maintaining a static catalog.
4. Add a provider documentation page with a minimal config.yaml example and links to the API key page and API docs.
5. Keep the first integration scoped to Chat Completions; no provider-specific payload transformation is required.

I can contribute the implementation and focused tests following the existing provider patterns.

Disclosure: I represent Modelsell.

Start by locating Continue's existing OpenAI-compatible Chat Completions implementation and provider patterns, then review how provider documentation and config.yaml examples are organized. Add the Modelsell provider with its default base URL, bearer-key setup, dynamic model discovery, and documentation links; focused tests should cover the provider defaults and model listing.