[WRONG BRANCH] fix(claude): preserve pre-output keepalives

[luvs01]

Motivation

• A recent change deferred Anthropic message_start until semantic output but unintentionally suppressed keepalive pings during the upstream pre-semantic silence, causing idle downstream connections to drop behind LBs/NATs.
• The intent of this change is to restore transport-level keepalive behavior so periodic pings (and upstream heartbeats) flow even before the first semantic frame arrives while still avoiding manufacturing a message_start before an initial error.

Description

• Allow the periodic ping timer to emit transport ping frames regardless of started by removing the !started gate in the interval callback and updating the comment to clarify transport-only semantics.
• Relay upstream response.heartbeat frames as transport ping frames even before Anthropic framing starts so upstream heartbeats keep the downstream connection alive.
• Add two regression tests in tests/claude-outbound.test.ts: idle keepalive pings flow before semantic output during upstream silence and upstream heartbeats remain transport-only before an initial error to cover the pre-semantic idle and pre-error heartbeat behaviors.

Testing

• Ran bun test tests/claude-outbound.test.ts and the targeted file's tests passed (40 passed, 0 failed).
• Ran bun run typecheck (bun x tsc --noEmit) which completed successfully.
• Attempted full bun run test; the full suite encountered unrelated, environment-dependent failures (e.g. combo-management timeouts and a few Docker/lab-specific expectations) that are outside the scope of this focused fix.

---

Codex Task

[coderabbitai]

Warning

Review limit reached

@luvs01, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 41 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 3e59b702-5ead-4147-afcd-058d13266563

📥 Commits

Reviewing files that changed from the base of the PR and between 6d881db and 390a5ba.

📒 Files selected for processing (2)

• src/claude/outbound.ts
• tests/claude-outbound.test.ts

---

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

[github-actions]

✅ Deterministic PR hygiene checks passed.

[github-actions]

⏳ DRAFT

• wrong target branch (main); retarget to dev.

What to do

• Retarget this PR to dev — all contributions go to dev.

Its title has been prefixed with [WRONG BRANCH].
This pull request was already a draft. Its draft status will be preserved after every issue above is resolved.