fix(mobile): keep review usable without its native diff view

[rynfar]

Opening a parsed diff can crash when the native diff view has not registered or is missing from the installed mobile binary. Fall back to its raw patch, hide the navigator that cannot scroll that fallback, and provide pull-to-refresh through the existing refresh handler.

Adopts pingdotgg/t3code#9828 (bc3dc2694b5d15613ef610d693d18ebd904b0d56) with Pylon's current review screen, theme, navigation, and connection handling. The native diff path remains available when registered. This is JavaScript-only and does not change the fingerprint policy or native dependency set.

Validation: mobile tsc --noEmit, scoped lint and the existing focused review availability, pane-selection and model tests pass. An integrated iPhone 17 Pro / iOS 26.3 check reproduced the null-component crash before the fix, rendered the raw patch after the fix, and loaded changed fixture contents with a native pull-to-refresh gesture. Restoring normal native resolution returned to the native diff surface. Temporary fault injection was removed; the integrated mobile package typecheck passed. Android shares this component but was not exercised in a device runtime. Before/after images and the refresh video are attached below. Web/desktop are unaffected; review behavior is independent of the provider and still uses the selected environment.

Before	After	After refresh

Native pull-to-refresh recording.

Model: GPT-6 Astra. Harness: Codex.

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	13.7 KiB	13.7 KiB	−3 B (−0.0%)	15.1 KiB	✅
Codex	Thread snapshot wire	7.1 KiB	7.1 KiB	−3 B (−0.0%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.6 KiB	6.6 KiB	0 B (0.0%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	56.5 KiB	56.5 KiB	0 B (0.0%)	66.4 KiB	✅
Codex	Live turn messages	8	8	0 (0.0%)	21	✅
Claude	Total thread wire	13.6 KiB	13.7 KiB	+51 B (+0.4%)	15.1 KiB	✅
Claude	Thread snapshot wire	7.1 KiB	7.1 KiB	+8 B (+0.1%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.5 KiB	6.6 KiB	+43 B (+0.6%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	57.3 KiB	57.3 KiB	+44 B (+0.1%)	66.4 KiB	✅
Claude	Live turn messages	7	8	+1 (+14.3%)	21	✅

Baseline: 983633a · PR result: 3a14944 · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 111.2 KiB
• Claude decoded thread snapshot: 111.9 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.