fix: SIP connect hang - arm timeout before await connect

[coygg]

Fixes agents stuck on WebRTC connecting forever. Timeout now races against connect(), and ensureRegistered recovers from stuck connecting state.

Summary by CodeRabbit

• Bug Fixes
  • Improved handling of connection timeouts to more reliably mark initialization complete and surface an error state when SIP connect fails.
  • Added better detection and logging for “stuck connecting” situations.
  • Triggered automatic recovery attempts when connection attempts exceed the allowed timeout, improving registration resilience.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
policyjar	Ready	Preview, Comment	Apr 3, 2026 10:36pm

[coderabbitai]

Caution

Review failed

The pull request is closed.

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro

Run ID: eca67ac3-7061-4c4f-ac74-d79f133144fe

📥 Commits

Reviewing files that changed from the base of the PR and between fdb2f68 and 4c0c93b.

📒 Files selected for processing (1)

• src/hooks/use-telnyx.ts

---

📝 Walkthrough

Walkthrough

Hook initialization now records connect start time and races client.connect() against a timeout. On timeout the hook marks initialization complete, sets connection status to error, clears registration, and rejects. ensureRegistered() now detects prolonged connecting state and triggers a manual reconnect.

Changes

Cohort / File(s)

Summary

Connection initialization & recovery src/hooks/use-telnyx.ts

Added connectStartTimeRef. Refactored init to set the start time before awaiting client.connect() and race it with a pre-armed timeout promise. Timeout path now logs "SIP connect timeout", calls markInitialized(), sets connectionStatus to 'error', sets isRegistered to false, and rejects with Error('SIP connect timeout'). ensureRegistered() now measures elapsed connect time when connectionStatusRef.current === 'connecting' and, if exceeded, logs a "stuck connecting" warning and calls triggerManualReconnect('connect-timeout').

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Poem

🐰 I stamped my paw at the ticking line,
A race of promises, snap!—not fine.
When Telnyx lingers past the chime,
I nudge reconnect, reset the time.
Hoppity hop—back in rhyme!

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title directly and clearly summarizes the main fix: addressing a SIP connect hang by arming a timeout before awaiting the connect call, which aligns with the core change in the changeset.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/sip-connect-hang

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@src/hooks/use-telnyx.ts`:
- Around line 744-753: The stuck-connecting branch in ensureRegistered detects a
connection hang (connectionStatusRef.current === 'connecting' and elapsed >
REGISTRATION_TIMEOUT_MS) but only calls runLifecycleRecovery('connect-timeout')
which merely schedules another ensureRegistered and will be deduped, so it never
forces a reconnect; change this to call
triggerManualReconnect('connect-timeout') (the same recovery used in other
failure paths) instead of or in addition to runLifecycleRecovery so the client
actually disconnects/reinits when the connect timeout is reached; update the
call site in ensureRegistered (and keep runLifecycleRecovery if you want logging
scheduling still) and ensure triggerManualReconnect is invoked with the
'connect-timeout' reason.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro

Run ID: 2e30ffe9-815a-4754-96e6-7181ec0192d9

📥 Commits

Reviewing files that changed from the base of the PR and between f78147c and fdb2f68.

📒 Files selected for processing (1)

• src/hooks/use-telnyx.ts

Findings addressed.