fix: always accept 'auto' in defaultMode settings schema

[luoxk]

Summary

• Fix SettingsSchema to always accept 'auto' as a valid permissions.defaultMode value, regardless of the TRANSCRIPT_CLASSIFIER feature flag state.

Problem

When TRANSCRIPT_CLASSIFIER is disabled (sidecar/desktop builds), defaultMode: 'auto' in ~/.claude/settings.json causes Zod validation to fail. This silently discards the entire user settings file, resulting in:

• enabledPlugins lost → all plugin skills unavailable → "Unknown skill" errors
• hooks, env, extraKnownMarketplaces all lost
• Only project-local .claude/settings.local.json is loaded

Users who configured "defaultMode": "auto" in the upstream CLI (where the flag is on) cannot use their settings in sidecar builds.

Fix

Replace the feature-flag-gated enum:

// Before
defaultMode: z.enum(
  feature('TRANSCRIPT_CLASSIFIER') ? PERMISSION_MODES : EXTERNAL_PERMISSION_MODES,
)

// After
defaultMode: z.enum([...EXTERNAL_PERMISSION_MODES, 'auto'])

The schema should accept all valid mode values for file compatibility. Runtime behavior for auto mode is separately gated by the feature flag — the schema just needs to not reject the file.

Test plan

• Verified: sidecar CLI subprocess with defaultMode: 'auto' in user settings now correctly loads enabledPlugins and all plugin skills
• Verified: SettingsSchema().safeParse() no longer returns success: false for files containing defaultMode: 'auto'

[github-actions]

PR quality triage

Changed areas: area:cli-core

CLI core policy: Blocked by policy until a maintainer applies allow-cli-core-change and approves the PR.

Missing-test policy: Blocked by policy until a maintainer applies allow-missing-tests or matching tests are added.

Coverage baseline policy: No coverage-baseline policy block detected.

CLI core files:

• src/utils/settings/types.ts

Coverage policy files:

• none

Expected checks:

• change-policy
• server-checks
• coverage-checks

Test coverage signals:

• BLOCKING unless allow-missing-tests is applied: Agent/runtime product files changed without a tools/utils test file in the PR.
• Agent/model runtime path changed: use mock/request-shape tests in PR and maintainer live-model smoke before release.

Risk notes:

• No special risk notes from changed paths.

Hard merge gates still come from GitHub Actions, not AI review.

Dosu handoff: Dosu can be used as the AI reviewer for risk explanation, missing-test prompts, and maintainer Q&A. If it does not comment automatically from the PR template, ask:

@dosubot review this PR for changed-area risk, missing tests, docs impact, desktop startup risk, and CLI core impact.