feat: Inception (Mercury 2) provider + API params & temperature UI adjustment

[jadeezomg]

Enjoying their diffusion based model a lot for quick questions, so I wanted to add native support. For testing they offer API access with 10m tokes just with sign-up. Have tested all options

AI Disclosure: Change was mostly done via Cursor

PR Description

Adds support for Inception Platform / Mercury 2 and aligns the settings UX with their API parameters.

Provider: inception

• New provider preset: base https://api.inceptionlabs.ai/v1, default model mercury-2.
• Uses dedicated inceptionRequest in AIApiAdapters.js (OpenAI-style chat completions + Mercury 2 fields).
• API keys: saved/session/env; also INCEPTION_API_KEY / DMS_INCEPTION_API_KEY.

Mercury 2 request body (from docs)

• reasoning_effort: instant | low | medium | high (dropdown, default medium).
• reasoning_summary: toggle (default on).
• reasoning_summary_wait: toggle (default off).
• temperature: clamped 0.5–1.0 in the adapter; max_tokens clamped 1–50000 (default 8192 when unset for Inception).
• stream: true unchanged (keeps Stop/cancel behavior).

Intentionally not exposed

• diffusing — not useful with the current chat UI (no meaningful step-by-step display).
• stop / stream_options.include_usage — removed to keep settings simple.

Small change: temperature display

• Settings temperature slider used 0.1 steps; 0.75 rounded to 0.8. Slider is now 0.01 steps (0–2.00) and label shows two decimals (e.g. 0.75).