fix(release): remove unused Expo patches

[rynfar]

Problem

The release smoke test regenerates pnpm-lock.yaml from the workspace manifests. A fresh SDK 57 resolution now selects expo-sharing@57.0.17 from the mobile app's ~57.0.16 range. It also selects expo-modules-core@57.0.15, whose ~57.0.7 dependency selects expo-modules-jsi@57.0.7. No workspace manifest pins the old patched versions, so pnpm rejects the stale expo-sharing@57.0.16 and expo-modules-jsi@57.0.6 patch registrations as unused.

Both replacement tarballs still lack Pylon's required fixes: the sharing package does not preserve Android share filenames/mixed text-and-file payloads, and the JSI package still marks its runtime scheduler constructors SWIFT_RETURNS_RETAINED under Xcode 26.3.

Fix

Retarget the two existing patches, without changing their semantics, to the versions selected by a clean lockfile resolution. Refresh only the related Expo lock entries. Unused-patch enforcement remains enabled, and every other patch is unchanged.

Evidence

• Exact base 22e8c2b616fa52cea9cebac6505e31a0740fdb4e: node scripts/release-smoke.ts reproduced ERR_PNPM_UNUSED_PATCH for both old Expo versions.
• vp install --lockfile-only --ignore-scripts — passed; Vite Plus reported that the lockfile passes supply-chain policies.
• node scripts/release-smoke.ts — passed (Release smoke checks passed.).
• vp run --filter @t3tools/mobile typecheck — passed.
• Six focused mobile sharing test files — 65 tests passed.
• Focused mobile sharing lint — 0 warnings and 0 errors.
• vp fmt --check pnpm-workspace.yaml pnpm-lock.yaml — passed.
• git diff --check HEAD^ — passed.

Model: GPT-5.6
Harness: Prime Agent

---

^{Need help on this PR? Tag @codesmith-bot with what you need. Autofix is disabled.}

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

ℹ️ The exact PR base did not have a successful artifact. Baseline uses the latest successful main measurement shown below.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	13.8 KiB	13.5 KiB	−235 B (−1.7%)	15.1 KiB	✅
Codex	Thread snapshot wire	6.9 KiB	6.9 KiB	−9 B (−0.1%)	7.3 KiB	✅
Codex	Live turn WebSocket wire	6.8 KiB	6.6 KiB	−226 B (−3.2%)	7.8 KiB	✅
Codex	Live turn WebSocket decoded	58.7 KiB	57.1 KiB	−1.5 KiB (−2.6%)	66.4 KiB	✅
Codex	Live turn messages	11	9	−2 (−18.2%)	21	✅
Claude	Total thread wire	13.8 KiB	13.5 KiB	−230 B (−1.6%)	15.1 KiB	✅
Claude	Thread snapshot wire	6.9 KiB	6.9 KiB	+2 B (+0.0%)	7.3 KiB	✅
Claude	Live turn WebSocket wire	6.8 KiB	6.6 KiB	−232 B (−3.3%)	7.8 KiB	✅
Claude	Live turn WebSocket decoded	59.5 KiB	58.0 KiB	−1.5 KiB (−2.5%)	66.4 KiB	✅
Claude	Live turn messages	11	9	−2 (−18.2%)	21	✅

Baseline: 03defac · PR result: 10e4718 · 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: 109.5 KiB
• Claude decoded thread snapshot: 110.2 KiB

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

[rynfar]

Superseded by #243, which landed the same reviewed patch retargets plus exact manifest pins. Its full CI and release smoke are green.