Assessment Models: Add Anthropic Claude

[vprashrex]

Is your feature request related to a problem?
The AI Assessment model picker currently only includes OpenAI and Google models, excluding Anthropic's Claude models. This limits users who want to utilize these models for their assessments despite existing support for model configuration in the pipeline.

Describe the solution you'd like

• Add an anthropic provider entry in app/lib/data/assessmentModels.ts for Claude models: claude-sonnet-4-6 and claude-haiku-4-5
• Define an Anthropic config preset similar to GEMINI_TEMPERATURE_CONFIG / GPT4_STYLE_CONFIG
• Surface the new provider in the UI, allowing users to select it alongside OpenAI and Gemini
• Update app/lib/types/assessment.ts if the provider union is mentioned there

Original issue

Describe the current behavior
The AI Assessment model picker (app/lib/data/assessmentModels.ts) currently only offers OpenAI and Google (Gemini) models. Users who want to run assessments with Anthropic's Claude models cannot select them, even though the assessment pipeline already supports per-provider model configuration (added for Gemini in #153).

Describe the enhancement you'd like

• Add an anthropic provider entry set to app/lib/data/assessmentModels.ts with the current Claude models: claude-sonnet-4-6 (Claude Sonnet 4.6), and claude-haiku-4-5 (Claude Haiku 4.5)
• Define an Anthropic config preset (mirroring GEMINI_TEMPERATURE_CONFIG / GPT4_STYLE_CONFIG) Surface the new provider in the assessment configuration UI (provider dropdown and model select) alongside OpenAI and Gemini
• Update app/lib/types/assessment.ts provider types if the provider union is enumerated there