feat(export): add NVIDIA opt-in detection

[meiiie]

Summary

• Add a main-process native export capability probe for NVIDIA CUDA availability.
• Surface an experimental NVIDIA CUDA switch in the MP4 Lightning export menu only when this Windows device reports a usable NVIDIA path.
• Require user opt-in before packaged/user-facing CUDA candidates run, while preserving the explicit lab env path for RECORDLY_EXPERIMENTAL_NVIDIA_CUDA_EXPORT=1.

Why

This carries forward the NVIDIA research from #504 without merging the broad draft branch. Users should not silently enter an experimental CUDA path just because the app is packaged; the app should detect the device, show an explicit opt-in, and continue to rely on the existing validation/fallback chain.

Platform Scope

• This PR intentionally enables the NVIDIA CUDA opt-in only for Windows because the current Recordly CUDA helper is packaged as a Windows .exe and the export route still falls back through the Windows GPU compositor.
• Linux NVIDIA/CUDA/NVENC is supported by NVIDIA at the platform level, but Recordly does not currently package or validate a Linux CUDA compositor helper in this slice.
• A future Linux NVIDIA PR should be separate and must add Linux helper packaging, Linux capability probing, packaged-app smoke coverage, and real-device QA before exposing any Linux UI switch.

References Checked

• Electron app.getGPUInfo('basic') for vendor/device detection: https://www.electronjs.org/docs/latest/api/app#appgetgpuinfoinfotype
• NVIDIA CUDA Linux prerequisites and GPU verification: https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html
• NVIDIA Video Codec SDK Windows/Linux support: https://developer.nvidia.com/video-codec-sdk
• NVIDIA FFmpeg GPU acceleration Windows/Linux requirements: https://docs.nvidia.com/video-technologies/video-codec-sdk/13.0/ffmpeg-with-nvidia-gpu/index.html
• NVIDIA NVML platform notes for deeper future probing: https://docs.nvidia.com/deploy/nvml-api/nvml-api-reference.html

Safety Notes

• No CUDA source, binary, helper manifest, or dev launcher changes in this PR.
• The switch is runtime/app-local, not saved into project files or presets.
• If CUDA is opted in and audio is present, the native path uses the shared mux validation guard instead of trusting inline audio.
• Explicit lab env usage is preserved for developer testing.

Verification

• npm test -- electron/ipc/export/native-video.test.ts
• npm test -- src/lib/exporter/modernVideoExporter.nativeStaticLayout.test.ts
• npx tsc --noEmit
• npx biome check --formatter-enabled=false electron/ipc/export/native-video.ts electron/ipc/export/native-video.test.ts electron/ipc/register/export.ts electron/preload.ts electron/electron-env.d.ts src/lib/exporter/types.ts src/lib/exporter/modernVideoExporter.ts src/components/video-editor/ExportSettingsMenu.tsx src/components/video-editor/VideoEditor.tsx
• git diff --check
• npx vite build --config vite.config.ts
• npm run smoke:electron-main-cjs

Not Run

• npm run smoke:packaged-binaries because this checkout does not currently have a packaged release/app.asar.unpacked directory.

Related to #504.

[coderabbitai]

📝 Walkthrough

Walkthrough

Adds an experimental NVIDIA CUDA native export path with user opt‑in, runtime capability probing via getNativeExportCapabilities(), refactors GPU probe behavior, updates skip‑reason/timeout logic, wires IPC/preload/exporter/UI, and adds tests.

Changes

NVIDIA CUDA Export Capability with User Opt-In

Layer / File(s)	Summary
Capability types and API contracts electron/electron-env.d.ts, electron/preload.ts, src/lib/exporter/types.ts	Declares renderer/preload/exporter capability types and exposes getNativeExportCapabilities(); extends nativeStaticLayoutExport and ExportConfig with experimentalNvidiaCudaExport option.
Native NVIDIA CUDA capability detection and eligibility refactor electron/ipc/export/native-video.ts	Refactors GPU probing to return null when probe unavailable; adds getNativeExportCapabilities() that checks wrapper/GPU inspectability; replaces packaged-auto-candidate semantics with explicit env and user experimentalNvidiaCudaExport opt-in; updates validated-fallback helpers, audio mux validation, stall timeout selection, and skip-logging metadata.
IPC handler registration electron/ipc/register/export.ts	Registers get-native-export-capabilities IPC handler that returns structured success/error responses from getNativeExportCapabilities().
Export pipeline integration src/lib/exporter/modernVideoExporter.ts	ModernVideoExporter forwards experimentalNvidiaCudaExport flag into nativeStaticLayoutExport.
VideoEditor capability detection and export decision logic src/components/video-editor/VideoEditor.tsx	Adds persisted opt-in state, queries getNativeExportCapabilities() on mount to set availability and auto-disables when unsupported, persists opt-in changes, forces modern pipeline on enable, and computes useExperimentalNvidiaCudaExport for MP4 export.
ExportSettingsMenu UI component src/components/video-editor/ExportSettingsMenu.tsx	Adds experimentalNvidiaCudaExport, onExperimentalNvidiaCudaExportChange, and nvidiaCudaExportAvailable props with defaults; conditionally renders MP4-only CUDA toggle when available.
Test coverage electron/ipc/export/native-video.test.ts	Adds getNativeExportCapabilities test import; updates getNvidiaCudaAutoStallTimeoutMs description; introduces getNativeExportCapabilities test suite; refactors getExperimentalNvidiaCudaExportSkipReason tests to cover user opt-in and env-disable cases.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Possibly related PRs

• webadderallorg/Recordly#559: Overlaps on getExperimentalNvidiaCudaExportSkipReason() gating logic and packaged auto-candidate vs validated behavior.
• webadderallorg/Recordly#556: Related native static-layout export routing using capability/skipReason decisions.
• webadderallorg/Recordly#555: Modifies modernVideoExporter native static-layout flow that also forwards CUDA-related options.

Suggested labels

Checked

Poem

🐰 A rabbit sniffs the GPU breeze,
"Opt-in for CUDA," says with ease.
Toggle on, check wrapper true,
Fall back neatly if probes won't do—
Hops, saves settings, exports with glee.

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title 'feat(export): add NVIDIA opt-in detection' clearly and concisely summarizes the main change—adding NVIDIA opt-in detection capability to the export system.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Description check	✅ Passed	PR description is comprehensive and well-structured, covering motivation, scope, platform constraints, safety considerations, and verification steps.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feature/nvidia-export-opt-in

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@src/components/video-editor/VideoEditor.tsx`:
- Around line 631-649: The effect that calls
window.electronAPI?.getNativeExportCapabilities?() must catch promise rejections
so an error doesn't become an unhandled rejection and so the UI falls back to a
deterministic state; update the useEffect (the async then handler around
getNativeExportCapabilities) to add a .catch handler (or use try/catch inside an
async IIFE) that treats any error as available = false, calls
setNvidiaCudaExportAvailable(false), resets
setExperimentalNvidiaCudaExportState(false) if needed, and optionally logs the
error, while still respecting the cancelled flag before updating state.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: a0ac908c-7ed9-4d86-a974-c70e3de02089

📥 Commits

Reviewing files that changed from the base of the PR and between 7353ba9 and c85dc8b.

📒 Files selected for processing (9)

• electron/electron-env.d.ts
• electron/ipc/export/native-video.test.ts
• electron/ipc/export/native-video.ts
• electron/ipc/register/export.ts
• electron/preload.ts
• src/components/video-editor/ExportSettingsMenu.tsx
• src/components/video-editor/VideoEditor.tsx
• src/lib/exporter/modernVideoExporter.ts
• src/lib/exporter/types.ts