Update Hebrew translation coverage

[MusiCode1]

Summary

• Add missing Hebrew translation keys for Speech navigation and deleted git changes.
• Fix Hebrew placeholder coverage for singular count strings and apply-patch file counts.
• Polish stale Hebrew wording in notifications and tool call labels.

Verification

• Hebrew catalog parity check: en=1027 he=1027 missing=0 extra=0 placeholderMismatch=0
• git diff --check -- packages/ui/src/lib/i18n/messages/he
• npm run typecheck --workspace @codenomad/ui fails on existing unrelated issues: missing virtua/solid, missing @tauri-apps/plugin-dialog, and implicit any parameters in virtual-follow-list.tsx.

[shantur]

@MusiCode1 - Seems like you need to update to latest dev

[pascalandr]

@CodeNomadBot

I want you to perform the gatekeeper checks for latest PR state

check for regressions
check for better possible implementations
Be the PR gatekeeper
Be ruthless about code quality

[codenomadbot]

Gatekeeper review for latest PR state.

Understood request: review PR #388 ruthlessly for regressions and better implementations. Assumption: scope is the current Hebrew translation-only diff, not unrelated existing i18n debt or local worktree changes.

Findings: none blocking.

What I checked:

• Reviewed the current PR diff against origin/dev: only four Hebrew catalog files changed: folderSelection.ts, instance.ts, settings.ts, and toolCall.ts.
• Verified the added/changed placeholders match English usage for the touched strings, including singular {count} strings and {path} in the save confirmation.
• Verified Hebrew catalog parity with English via a local parser: en=1374, he=1374, missing=0, extra=0, placeholderMismatch=0.
• Ran git diff --check -- packages/ui/src/lib/i18n/messages/he: passed with no whitespace errors.
• Ran npm run typecheck --workspace @codenomad/ui: failed, but not on this translation diff. The local workspace is missing broad dependency/type resolution (solid-js, @opencode-ai/sdk/v2, Tauri plugins, Node test types, etc.), so this run is not useful as a PR regression signal in the current checkout.

Quality notes:

• The changes are appropriately minimal and confined to Hebrew i18n coverage.
• The implementation matches the existing catalog structure; no new abstraction is warranted for this PR.
• Non-blocking recommendation: add a committed i18n parity/placeholder check script or CI job. This PR proves parity manually, but the repo currently relies on ad hoc checks, which makes future locale regressions easy to miss.

Verdict: approve from a gatekeeper/code-quality perspective for the current translation-only scope.

--
Yours,
CodeNomadBot

[pascalandr]

@MusiCode1
Thanks !

[ghost]

PR builds are available as GitHub Actions artifacts:

https://github.com/NeuralNomadsAI/CodeNomad/actions/runs/26420295880

Artifacts expire in 7 days.
Artifacts:

• pr-388-b3b0111b4e5c79e2eef591d1a5055efb2fcf5f7d-tauri-macos
• pr-388-b3b0111b4e5c79e2eef591d1a5055efb2fcf5f7d-tauri-linux
• pr-388-b3b0111b4e5c79e2eef591d1a5055efb2fcf5f7d-tauri-windows
• pr-388-b3b0111b4e5c79e2eef591d1a5055efb2fcf5f7d-electron-macos
• pr-388-b3b0111b4e5c79e2eef591d1a5055efb2fcf5f7d-tauri-macos-arm64
• pr-388-b3b0111b4e5c79e2eef591d1a5055efb2fcf5f7d-electron-windows
• pr-388-b3b0111b4e5c79e2eef591d1a5055efb2fcf5f7d-electron-linux