Best possible solution:

Land the narrow NodeAppModel failed-start cleanup after the branch refreshes native i18n inventory, supplies redacted real iOS proof for failed and successful PTT paths, and keeps existing stop/cancel resume behavior.

Do we have a high-confidence way to reproduce the issue?

Yes from source inspection, but not from an executed iOS run: current main suspends Voice Wake before beginPushToTalk(), and the offline gateway guard throws before stop/cancel cleanup can resume it.

Is this the best way to solve the issue?

Yes, conditionally: NodeAppModel is the right coordination point between Voice Wake and PTT, and the PR mirrors the existing pttOnce/camera defer pattern for failed starts. It still needs inventory sync and real iOS proof before it is merge-ready.

Full review comments:

• [P3] Update the native i18n inventory — apps/ios/Tests/NodeAppModelInvokeTests.swift:695
  Late note from this same-head re-review: the current native-i18n CI job fails with inventory drift and asks for pnpm native:i18n:sync plus a committed apps/.i18n/native-source.json update. The new Swift literals in this test are part of the PR diff, so the branch cannot pass required checks until the inventory is refreshed.
  Confidence: 0.92
  Late finding: first raised on code an earlier review cycle already covered.

Overall correctness: patch is correct
Overall confidence: 0.82

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning high; reviewed against 706443c79b1c.

Label changes:

• add rating: 🧂 unranked krab: Overall readiness is 🧂 unranked krab; proof is 🧂 unranked krab and patch quality is 🦐 gold shrimp.
• remove rating: 🦪 silver shellfish: Current PR rating is rating: 🧂 unranked krab, so this older rating label is no longer current.

Label justifications:

• P2: This is a bounded iOS Voice Wake/Talk recovery bug with user-visible impact, but it is limited to the failed PTT start path and has no data-loss or security signal.
• rating: 🧂 unranked krab: Overall readiness is 🧂 unranked krab; proof is 🧂 unranked krab and patch quality is 🦐 gold shrimp.
• status: 📣 needs proof: The PR needs real behavior proof before ClawSweeper can clear the contributor ask. Needs real behavior proof before merge: The PR body lists diff checks, autoreview, and a blocked local Xcode attempt, but no live iOS/PTT log, terminal output, screenshot, recording, or artifact shows the after-fix behavior. After adding proof, update the PR body; ClawSweeper should re-review automatically. If it does not, the PR author or someone with repository write access can comment @clawsweeper re-review.

PR surface:

Other +28. Total +28 across 3 files.

Acceptance criteria:

• [P1] pnpm native:i18n:check.
• [P1] pnpm ios:gen.
• [P1] xcodebuild test -project apps/ios/OpenClaw.xcodeproj -scheme OpenClaw -destination 'platform=iOS Simulator,name=iPhone 17,OS=26.5' -only-testing:OpenClawTests/NodeAppModelInvokeTests.

What I checked:

• current main gap: Current main suspends Voice Wake before beginPushToTalk() in talk.pttStart, but only pttStop and pttCancel resume it; a throw from the start path leaves the external-audio suspension uncleared. (apps/ios/Sources/Model/NodeAppModel.swift:1698, 706443c79b1c)
• throw path: beginPushToTalk() throws Gateway not connected before capture starts when gatewayConnected is false, which matches the PR's focused failure path. (apps/ios/Sources/Voice/TalkModeManager.swift:648, 706443c79b1c)
• resume contract: resumeAfterExternalAudioCapture(wasSuspended:) clears isSuspendedForExternalAudio and schedules Voice Wake restart only when the caller reports a real suspension. (apps/ios/Sources/Voice/VoiceWakeManager.swift:299, 706443c79b1c)
• PR patch shape: The PR head adds the didStart/defer failed-start cleanup around beginPushToTalk() while preserving successful-start cleanup for stop/cancel. (apps/ios/Sources/Model/NodeAppModel.swift:1699, 2efc949c6f23)
• required check failure: The current head's native-i18n CI job fails with native app i18n inventory drift detected and instructs running pnpm native:i18n:sync and committing apps/.i18n/native-source.json. (apps/.i18n/native-source.json, 2efc949c6f23)
• related PR check: fix: prevent Voice Wake crash after Talk audio capture #99137 is merged and hardens adjacent Voice Wake external-capture lifecycle paths, but it does not add failed-start cleanup to NodeAppModel's pttStart path on current main. (apps/ios/Sources/Voice/VoiceWakeManager.swift:287, 2f707f981152)

Likely related people:

• mbelinky: Authored the iOS push-to-talk node commands, PTT once/cancel handling, and the original Voice Wake pause during PTT flow that this cleanup path builds on. (role: introduced behavior; confidence: high; commits: 9f101d3a9aae, 1a48bce2947e, 17b18971f17c; files: apps/ios/Sources/Model/NodeAppModel.swift, apps/ios/Sources/Voice/TalkModeManager.swift)
• PollyBot13: Authored the recent merged Voice Wake external-capture and restart hardening that shares the same audio lifecycle surface. (role: recent area contributor; confidence: high; commits: 2f707f981152, 5bc616588f17, 1c6cf77d45eb; files: apps/ios/Sources/Voice/VoiceWakeManager.swift, apps/ios/Tests/VoiceWakeManagerSuppressionTests.swift)
• steipete: Merged the adjacent Voice Wake hardening PR and authored a recent follow-up removing Voice Wake taps after failed starts. (role: recent reviewer, merger, and adjacent contributor; confidence: high; commits: 2f707f981152, 933a88cdc66d, 16a5f0b006; files: apps/ios/Sources/Voice/VoiceWakeManager.swift, apps/ios/Sources/Voice/TalkModeManager.swift)

• 🦀 challenger crab: rare, exceptional readiness with strong proof, clean implementation, and convincing validation.
• 🦞 diamond lobster: very strong readiness with only minor maintainer review expected.
• 🐚 platinum hermit: good normal PR, likely mergeable with ordinary maintainer review.
• 🦐 gold shrimp: useful signal, but proof or patch confidence is still limited.
• 🦪 silver shellfish: thin signal; proof, validation, or implementation needs work.
• 🧂 unranked krab: not merge-ready because proof is missing/unusable or there are serious correctness or safety concerns.
• 🌊 off-meta tidepool: rating does not apply to this item.

Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics.

• ClawSweeper keeps one durable marker-backed review comment per issue or PR.
• Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
• A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
• PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
• Maintainers can also comment @clawsweeper review to request a fresh review only.
• Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
• Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
• Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.