[rschlaefli]

Ideas generation offered model selection but no writing-style choice. This adds the existing Default/DF selector and passes the selected guide into the original structured generation request.

The writing assistant and Ideas share one selector. Default retains existing behavior; invalid or unavailable styles fail before generation. Results remain title/body idea briefs, with no schema or provider changes.

Screenshots

Local synthetic application: English desktop (1280×900) and German phone (390×844). Captured at 8e29b85c; the final candidate changes only a dead test guard, with identical rendered application source. Provider responses are synthetic; these images do not demonstrate real model writing quality.

State	Desktop · English	Phone · German
Model and style selection
Generated idea brief

Validation: 108 focused backend tests; two Ideas browser cases and four existing assistant browser cases; Pint, frontend formatting/lint, TypeScript and the asset build. Browser checks cover model/style pairing, regeneration, switching to Default, save and reopening/reset. Independent source review is complete; its one advisory test cleanup is applied.

The focused Opengrep scan found no rule issues. Its parser rejects readonly constructor promotion; that class also passed a scan-only copy omitting the modifier, native PHP checks and direct source review.

Tests and lint passed on e5cc9ef6, including PostgreSQL, MySQL and the full browser suite. Paid model quality and deployment are outside this verification. Implementation plan.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota.

An organization owner can choose the repository organization as the billing source for reviews requested by members with a paid Copilot license. Learn about Copilot code review billing.