Problem

When the gateway restarts, loadActiveCalls() reloads non-terminal calls from calls.jsonl. However, these calls may have already ended (e.g., Twilio timed them out, or webhook couldn't reach local URL) and are now stale. This causes the concurrent call limit to be reached with phantom calls.

Solution

• Add getCallStatus() method to VoiceCallProvider interface
• Implement for all providers (Twilio, Plivo, Telnyx, Mock) using guardedJsonApiRequest
• On initialize(), verify each persisted non-terminal call with the provider before restoring
• Skip calls the provider reports as terminal (completed, failed, etc.)
• Skip calls older than maxDurationSeconds as a time-based fallback
• Skip calls without providerCallId (can't verify)

Fixes from review feedback

• P0 — Restart max-duration timers: Restored calls in answered+ state now get their hangup timers restarted with remaining time calculated from answeredAt
• P1 — Transient error handling: Provider lookup failures (network, 5xx) return isUnknown: true → call is kept with timer fallback. Only 404/known-terminal statuses drop the call
• P1 — Telnyx is_alive safety: Missing is_alive field treated as unknown (not terminal)
• Security: All provider GET calls use guardedJsonApiRequest (SSRF-guarded)

Testing

• 6 new unit tests covering verification scenarios (terminal, active, unknown, expired, no providerCallId, verification failure)
• All 84 voice-call tests pass
• pnpm tsgo + pnpm check clean

Test plan

1. Initiate a call
2. Kill the gateway before the call ends
3. Restart the gateway
4. Verify that stale calls are logged as skipped (provider status check)
5. Verify that new calls can be initiated