fix(transcription): honor configured default provider

[turboazot]

Summary

Describe this PR in 2-5 bullets:

• Base branch target (master for all contributions): master
• Problem: Telegram and WhatsApp voice-message transcription still routed through the legacy transcribe_audio() helper, which always instantiated GroqProvider and ignored transcription.default_provider
• Why it matters: OpenAI and other configured STT providers could never be used from these channel paths, causing runtime failures even when [transcription.openai] was configured correctly
• What changed: transcribe_audio() now dispatches by transcription.default_provider across the supported STT backends and adds a focused regression test for the OpenAI path
• What did not change (scope boundary): No schema changes, no provider implementation changes, no channel UX changes, no local deployment config included

Label Snapshot (required)

• Risk label (risk: low|medium|high): risk: medium
• Size label (size: XS|S|M|L|XL, auto-managed/read-only): size: S
• Scope labels (core|agent|channel|config|cron|daemon|doctor|gateway|health|heartbeat|integration|memory|observability|onboard|provider|runtime|security|service|skillforge|skills|tool|tunnel|docs|dependencies|ci|tests|scripts|dev, comma-separated): channel,tests
• Module labels (<module>: <component>, for example channel: telegram, provider: kimi, tool: shell): channel: telegram, channel: whatsapp_web
• Contributor tier label (trusted contributor|experienced contributor|principal contributor|distinguished contributor, auto-managed/read-only; author merged PRs >=5/10/20/50): auto-managed
• If any auto-label is incorrect, note requested correction: None

Change Metadata

• Change type (bug|feature|refactor|docs|security|chore): bug
• Primary scope (runtime|provider|channel|memory|security|ci|docs|multi): channel

Linked Issue

• Closes # N/A
• Related # N/A
• Depends on # N/A
• Supersedes # N/A

Supersede Attribution (required when Supersedes # is used)

• Superseded PRs + authors (#<pr> by @<author>, one per line): N/A
• Integrated scope by source PR (what was materially carried forward): N/A
• Co-authored-by trailers added for materially incorporated contributors? (Yes/No): No
• If No, explain why (for example: inspiration-only, no direct code/design carry-over): No superseded PR in this change
• Trailer format check (separate lines, no escaped \n): (Pass/Fail): Pass

Validation Evidence (required)

Commands and result summary:

cargo fmt --all -- --check
cargo clippy --all-targets -- -D warnings
cargo test

• cargo fmt --all -- --check: Pass
• cargo test openai_default_provider_uses_openai_config --lib: Pass
• cargo test manager_default_provider_from_config --lib: Pass
• Evidence provided (test/log/trace/screenshot/perf): focused unit-test output plus local runtime diagnosis showing the old Groq-only dispatch path
• If any command is intentionally skipped, explain why: cargo clippy --all-targets -- -D warnings and full cargo test were not run in this turn; focused regression coverage was used to validate the targeted bug fix before local release-fast rebuild/restart

Security Impact (required)

• New permissions/capabilities? (Yes/No): No
• New external network calls? (Yes/No): No
• Secrets/tokens handling changed? (Yes/No): No
• File system access scope changed? (Yes/No): No
• If any Yes, describe risk and mitigation: N/A

Privacy and Data Hygiene (required)

• Data-hygiene status (pass|needs-follow-up): pass
• Redaction/anonymization notes: No personal data or secrets added to repo changes
• Neutral wording confirmation (use ZeroClaw/project-native labels if identity-like wording is needed): Confirmed

Compatibility / Migration

• Backward compatible? (Yes/No): Yes
• Config/env changes? (Yes/No): No
• Migration needed? (Yes/No): No
• If yes, exact upgrade steps: N/A

i18n Follow-Through (required when docs or user-facing wording changes)

• i18n follow-through triggered? (Yes/No): No
• If Yes, locale navigation parity updated in README*, docs/README*, and docs/SUMMARY.md for supported locales (en, zh-CN, ja, ru, fr, vi)? (Yes/No): N/A
• If Yes, localized runtime-contract docs updated where equivalents exist (minimum for fr/vi: commands-reference, config-reference, troubleshooting)? (Yes/No/N.A.): N/A
• If Yes, Vietnamese canonical docs under docs/i18n/vi/** synced and compatibility shims under docs/*.vi.md validated? (Yes/No/N.A.): N/A
• If any No/N.A., link follow-up issue/PR and explain scope decision: N/A

Human Verification (required)

What was personally validated beyond CI:

• Verified scenarios: OpenAI default-provider path now returns the OpenAI-specific missing-key/config error rather than falling back to Groq; manager default-provider test still passes
• Edge cases checked: unsupported provider names still error explicitly; missing nested provider config produces provider-specific guidance
• What was not verified: full cargo test/clippy run and end-to-end Telegram voice-message transcription against a live OpenAI STT request in upstream CI context

Side Effects / Blast Radius (required)

• Affected subsystems/workflows: voice-message transcription entrypoints used by Telegram and WhatsApp Web
• Potential unintended effects: provider-specific error strings may surface earlier for non-Groq defaults; additional providers now follow default_provider via the legacy helper path
• Guardrails/monitoring for early detection: focused unit tests; runtime logs already surface Voice transcription failed with provider-specific errors

Agent Collaboration Notes (recommended)

• Agent tools used (if any): Codex terminal tools, local git, local systemd for repro only
• Workflow/plan summary (if any): reproduced runtime failure, traced it to the legacy wrapper, patched provider dispatch, added regression coverage, validated focused tests
• Verification focus: transcription routing for non-Groq defaults
• Confirmation: naming + architecture boundaries followed (AGENTS.md + CONTRIBUTING.md): Yes

Rollback Plan (required)

• Fast rollback command/path: revert commit 3d0872c1 or revert this PR
• Feature flags or config toggles (if any): None
• Observable failure symptoms: voice-message transcription regresses to provider misrouting or provider-specific configuration errors

Risks and Mitigations

List real risks in this PR (or write None).

• Risk: legacy helper now contains provider dispatch logic parallel to TranscriptionManager
  • Mitigation: scope is limited to preserving current public helper behavior; regression test added for the OpenAI default-provider path

[theonlyhennygod]

LGTM. Properly dispatches on the configured default_provider instead of always using Groq, with clear error messages and a test covering the openai path.