[juliusmarminge]

SimlockDeviceHost.renewLoop renewed every 2 minutes with a hard-coded { ttlMs: 10 minutes }. Simlock refuses (400) a renewal whose ttlMs is above the gateway's lease.maxTtlMs instead of shortening it, so on a gateway capped below 10 minutes every renewal failed and a watched lease expired after its original TTL. The lease is opened without a TTL, so it runs for the gateway's default length. The renewal now sends no ttlMs, and Simlock keeps that length, which a gateway always accepts. The unused LEASE_TTL_MS constant is removed and the comment above RENEW_INTERVAL says why no ttlMs is sent. No parser or regex added. The repro test on factory/repro-35982 is kept unchanged.

Fixes #35982

Reproduction on main

On origin/main a621be8, in apps/server/src/device/SimlockDeviceHost.test.ts, the existing fakePool gained a maxTtlMs option. Its fake gateway sets the lease deadline to open+maxTtlMs, answers 400 BAD_REQUEST to a renew whose ttlMs is above the cap, answers 404 to a renew after the deadline, and otherwise moves the deadline forward. New test: open a device on a host with maxLeaseMinutes 60 against a gateway capped at 5 minutes, then TestClock.adjust('6 minutes'). On main the lease is gone from ready.devices.list: the renewals at 2 and 4 minutes ask for ttlMs 600000 and get 400, and the one at 6 minutes gets 404, so renewLoop forgets the lease. Changing the renew body to {} makes all 13 tests in the file pass, which confirms the TTL in the request is the cause. That change was reverted and not committed. This ran the real SimlockDeviceHost adapter against a fake gateway, not a live Simlock deployment.

Evidence

Configure a 60-minute host against a gateway fixture capped at five minutes. Open a device through the real server RPC, subscribe to device state, and observe for 6.5 real minutes.

Before	After
a621be8d15	501c45a39f
Transcript playback, 15x	Transcript playback, 15x
Raw RPC output	Raw RPC output

These are captured backend transcripts rendered for inspection, not client screenshots. Both commits ran unmodified servers with the real pinned agent-device CLI against a synthetic HTTP gateway. Base returned 400/400/404 and dropped the lease; head returned 200/200/200 and retained it. Explicit close succeeded on head.

Reproduction driver, gateway fixture, run instructions, playback renderer.

Not shown:

• Live Simlock deployment and actual simulator deletion were not exercised: the gateway and device responses were synthetic.
• Device panel rendering and video continuity were not captured; observation used the real server's device-state subscription and device.list RPC.
• The configured sixty-minute maximum-age release was not observed; each comparison ran for 6.5 minutes.

Tests

vp test run src/device/SimlockDeviceHost.test.ts (apps/server): without the fix 1 failed / 12 passed (the new 'keeps a watched lease renewed when the gateway caps lease TTLs'); with the fix 13/13 passed. vp run check:changed: format, lint, knip, graphql, typecheck (3 packages), tests (2 files, 15 tests) all pass.

Not in this change

• RENEW_INTERVAL is a fixed 2 minutes; a gateway whose lease TTL is under 2 minutes would still let a watched lease expire between renewals. Fixing that would mean reading the lease's TTL and scaling the interval, a separate change.

Who did the work

• Written by Claude Opus 5.5, in Claude Code
• Reviewed by GPT-6.1 Sol, in Codex, and by Claude Opus 5.5, each on its own
• Proof captured by GPT-6.1 Sol and checked by Claude Opus 5.5

Run by the t3code factory in T3 Code.

[github-actions]

Thread transfer impact

✅ Thread transfer remains within every enforced ceiling.

Provider	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	Total thread wire	5.0 KiB	5.0 KiB	0 B (0.0%)	6.8 KiB	✅
Codex	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Codex	Live turn WebSocket wire	1.2 KiB	1.2 KiB	0 B (0.0%)	2.0 KiB	✅
Codex	Live turn WebSocket decoded	21.5 KiB	21.5 KiB	0 B (0.0%)	24.4 KiB	✅
Codex	Live turn messages	1	1	0 (0.0%)	8	✅
Claude	Total thread wire	5.0 KiB	5.0 KiB	0 B (0.0%)	6.8 KiB	✅
Claude	Thread snapshot wire	3.8 KiB	3.8 KiB	0 B (0.0%)	4.9 KiB	✅
Claude	Live turn WebSocket wire	1.2 KiB	1.2 KiB	0 B (0.0%)	2.0 KiB	✅
Claude	Live turn WebSocket decoded	21.9 KiB	21.9 KiB	0 B (0.0%)	24.4 KiB	✅
Claude	Live turn messages	1	1	0 (0.0%)	8	✅

Other scenarios

All 64 scenario measurements match main.

All 64 scenario measurements

Provider	Scenario	Metric	Main baseline	This PR	Impact	PR ceiling
Codex	short turn resume	WebSocket decoded	10.5 KiB	10.5 KiB	0 B (0.0%)	12.7 KiB	✅
Codex	short turn resume	WebSocket messages	2	2	0 (0.0%)	2	✅
Codex	long turn running, cold open	HTTP decoded	281.7 KiB	281.7 KiB	0 B (0.0%)	302.7 KiB	✅
Codex	long turn running, cold open	rows	75	75	0 (0.0%)	75	✅
Codex	long turn running, history page	HTTP decoded	92.4 KiB	92.4 KiB	0 B (0.0%)	107.4 KiB	✅
Codex	long turn running, history page	rows	75	75	0 (0.0%)	75	✅
Codex	long turn finished, cold open	HTTP decoded	381.5 KiB	381.5 KiB	0 B (0.0%)	410.2 KiB	✅
Codex	long turn finished, cold open	rows	151	151	0 (0.0%)	151	✅
Codex	long turn finished, socket cold open	WebSocket decoded	381.6 KiB	381.6 KiB	0 B (0.0%)	410.2 KiB	✅
Codex	long turn finished, socket cold open	WebSocket messages	2	2	0 (0.0%)	2	✅
Codex	long turn finished, socket cold open	rows	151	151	0 (0.0%)	151	✅
Codex	long turn finished, work page	HTTP decoded	186.2 KiB	186.2 KiB	0 B (0.0%)	200.2 KiB	✅
Codex	long turn finished, work page	rows	150	150	0 (0.0%)	150	✅
Codex	idle thread resume	WebSocket decoded	65 B	65 B	0 B (0.0%)	200 B	✅
Codex	idle thread resume	WebSocket messages	1	1	0 (0.0%)	1	✅
Codex	fleet shell snapshot	HTTP decoded	42.4 KiB	42.4 KiB	0 B (0.0%)	53.7 KiB	✅
Codex	fleet shell snapshot	rows	27	27	0 (0.0%)	27	✅
Codex	fleet shell subscribe	WebSocket decoded	65 B	65 B	0 B (0.0%)	200 B	✅
Codex	fleet shell subscribe	WebSocket messages	1	1	0 (0.0%)	1	✅
Codex	fleet shell resume check	HTTP decoded	18 B	18 B	0 B (0.0%)	100 B	✅
Codex	fleet shell resume check	rows	0	0	0	0	✅
Codex	stale shell reload	HTTP decoded	42.4 KiB	42.4 KiB	0 B (0.0%)	53.7 KiB	✅
Codex	stale shell reload	rows	27	27	0 (0.0%)	27	✅
Codex	stale shell resume	WebSocket decoded	140 B	140 B	0 B (0.0%)	400 B	✅
Codex	stale shell resume	WebSocket messages	2	2	0 (0.0%)	2	✅
Codex	stale todo list cold open	HTTP decoded	72.4 KiB	72.4 KiB	0 B (0.0%)	83.0 KiB	✅
Codex	stale todo list cold open	rows	75	75	0 (0.0%)	75	✅
Codex	stale todo list cold open	nodes	2	2	0 (0.0%)	4	✅
Codex	stale todo list cold open	subagents	2	2	0 (0.0%)	4	✅
Codex	stale todo list cold open	plans	1	1	0 (0.0%)	2	✅
Codex	stale todo list cold open	providerTurns	1	1	0 (0.0%)	2	✅
Codex	stale todo list cold open	checkpoints	1	1	0 (0.0%)	2	✅
Claude	short turn resume	WebSocket decoded	10.6 KiB	10.6 KiB	0 B (0.0%)	12.7 KiB	✅
Claude	short turn resume	WebSocket messages	2	2	0 (0.0%)	2	✅
Claude	long turn running, cold open	HTTP decoded	283.3 KiB	283.3 KiB	0 B (0.0%)	302.7 KiB	✅
Claude	long turn running, cold open	rows	75	75	0 (0.0%)	75	✅
Claude	long turn running, history page	HTTP decoded	93.4 KiB	93.4 KiB	0 B (0.0%)	107.4 KiB	✅
Claude	long turn running, history page	rows	75	75	0 (0.0%)	75	✅
Claude	long turn finished, cold open	HTTP decoded	384.2 KiB	384.2 KiB	0 B (0.0%)	410.2 KiB	✅
Claude	long turn finished, cold open	rows	151	151	0 (0.0%)	151	✅
Claude	long turn finished, socket cold open	WebSocket decoded	384.3 KiB	384.3 KiB	0 B (0.0%)	410.2 KiB	✅
Claude	long turn finished, socket cold open	WebSocket messages	2	2	0 (0.0%)	2	✅
Claude	long turn finished, socket cold open	rows	151	151	0 (0.0%)	151	✅
Claude	long turn finished, work page	HTTP decoded	188.1 KiB	188.1 KiB	0 B (0.0%)	200.2 KiB	✅
Claude	long turn finished, work page	rows	150	150	0 (0.0%)	150	✅
Claude	idle thread resume	WebSocket decoded	66 B	66 B	0 B (0.0%)	200 B	✅
Claude	idle thread resume	WebSocket messages	1	1	0 (0.0%)	1	✅
Claude	fleet shell snapshot	HTTP decoded	43.0 KiB	43.0 KiB	0 B (0.0%)	53.7 KiB	✅
Claude	fleet shell snapshot	rows	27	27	0 (0.0%)	27	✅
Claude	fleet shell subscribe	WebSocket decoded	66 B	66 B	0 B (0.0%)	200 B	✅
Claude	fleet shell subscribe	WebSocket messages	1	1	0 (0.0%)	1	✅
Claude	fleet shell resume check	HTTP decoded	18 B	18 B	0 B (0.0%)	100 B	✅
Claude	fleet shell resume check	rows	0	0	0	0	✅
Claude	stale shell reload	HTTP decoded	43.0 KiB	43.0 KiB	0 B (0.0%)	53.7 KiB	✅
Claude	stale shell reload	rows	27	27	0 (0.0%)	27	✅
Claude	stale shell resume	WebSocket decoded	140 B	140 B	0 B (0.0%)	400 B	✅
Claude	stale shell resume	WebSocket messages	2	2	0 (0.0%)	2	✅
Claude	stale todo list cold open	HTTP decoded	73.4 KiB	73.4 KiB	0 B (0.0%)	83.0 KiB	✅
Claude	stale todo list cold open	rows	75	75	0 (0.0%)	75	✅
Claude	stale todo list cold open	nodes	2	2	0 (0.0%)	4	✅
Claude	stale todo list cold open	subagents	2	2	0 (0.0%)	4	✅
Claude	stale todo list cold open	plans	1	1	0 (0.0%)	2	✅
Claude	stale todo list cold open	providerTurns	1	1	0 (0.0%)	2	✅
Claude	stale todo list cold open	checkpoints	1	1	0 (0.0%)	2	✅

Baseline: 43ce395 · PR result: 501c45a · Source CI: success

Scenario and decoded snapshot size

10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.

• Codex decoded thread snapshot: 108.5 KiB
• Claude decoded thread snapshot: 108.8 KiB

Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed.

[macroscopeapp]

Approvability

Verdict: Approved at 501c45a

Macroscope's review found this PR approvable — Renewals now preserve the lease’s gateway-assigned duration instead of requesting ten minutes, avoiding rejection by gateways with lower TTL caps. Renewal frequency, maximum lease age, and release behavior remain unchanged.

^{You can add or adjust custom eligibility rules. Learn more.}

[coderabbitai]

No actionable comments were generated in the recent review.

Walkthrough

The Simlock host no longer sends a TTL value when renewing a lease. The fake gateway test setup now models a five-minute TTL cap and verifies that a watched lease remains listed after six minutes.

Changes: Simlock lease renewal

Layer / File(s)	Summary
Renewal behavior and TTL-cap test apps/server/src/device/SimlockDeviceHost.ts, apps/server/src/device/SimlockDeviceHost.test.ts	The host omits ttlMs from renewal requests. The fake gateway models lease expiry and rejects TTL requests above its configured cap. A test checks that a watched lease remains listed after six minutes with a five-minute cap.

Estimated code review effort: 2 (Simple) | ~10 minutes

Change: Bug fix · Severity of issue fixed: Medium

Possibly related PRs

• pingdotgg/t3code#25259: Introduced the Simlock device host and its lease-renewal path.

Suggested reviewers: macroscopeapp

Merge Risk: Minimal · up to 501c4

Watched Simlock leases should now stay renewed on gateways that cap lease TTLs below ten minutes. The change is small and covered by a regression test. Gateways with TTLs under two minutes may still let leases expire, which is acknowledged as separate work.

Pre-merge checks: 4

Check name	Status	Explanation
Title check		The title names the main change: Simlock lease renewals no longer send a TTL that exceeds the gateway cap. It follows the conventional commit format and matches the description and the linked issue. I…
Linked Issues check		Issue #35982 requires that watched devices stay leased without requesting a TTL the gateway rejects. The diff in apps/server/src/device/SimlockDeviceHost.ts removes the { ttlMs: LEASE_TTL_MS } body fr…
Out of Scope Changes check		All changes relate to the linked issue. The source change removes the ttlMs renewal payload. It also removes the now-unused LEASE_TTL_MS constant and updates the RENEW_INTERVAL comment to explain why …
Description check		The description follows the template closely. It states the problem (a hard-coded 10-minute ttlMs makes renewals fail on gateways capped below 10 minutes, so watched leases expire), the change (renewa…

Full details: Title check

Explanation

The title names the main change: Simlock lease renewals no longer send a TTL that exceeds the gateway cap. It follows the conventional commit format and matches the description and the linked issue. It is a fix, and the code changes match the title.

Full details: Linked Issues check

Explanation

Issue #35982 requires that watched devices stay leased without requesting a TTL the gateway rejects. The diff in apps/server/src/device/SimlockDeviceHost.ts removes the { ttlMs: LEASE_TTL_MS } body from the renew call in renewLoop. Renewals now send no ttlMs, so Simlock keeps the lease's own length, and a gateway with lease.maxTtlMs=300000 no longer receives a rejected 600000 ms request. The new test 'keeps a watched lease renewed when the gateway caps lease TTLs' in SimlockDeviceHost.test.ts models the 5-minute cap and checks that the lease stays listed at 6 minutes. The PR description reports that this test fails without the fix and passes with it. The renewal-loop stop conditions and maxLeaseMinutes handling are not changed by this diff. The max-age bound therefore relies on existing code that this review did not re-read. The issue's scope is a local gateway simulation, and the PR states that scope. No coding requirement is unmet.

Full details: Out of Scope Changes check

Explanation

All changes relate to the linked issue. The source change removes the ttlMs renewal payload. It also removes the now-unused LEASE_TTL_MS constant and updates the RENEW_INTERVAL comment to explain why no ttlMs is sent. The test file adds a maxTtlMs option to fakePool and a regression test for the capped-gateway case. These support the fix. The PR does not change the fixed two-minute renewal interval, which the PR description states is a separate change. No unrelated change was found in the reviewed diff.

Full details: Description check

Explanation

The description follows the template closely. It states the problem (a hard-coded 10-minute ttlMs makes renewals fail on gateways capped below 10 minutes, so watched leases expire), the change (renewals omit ttlMs and the unused LEASE_TTL_MS constant is removed), and the scope (a bug fix tied to issue #35982, with the linked issue). Verification is detailed: the new test fails without the fix and passes with it (13/13), the repo check suite passes, and the proof limits are stated. The Problem heading is present, and the description names the author and tooling as the template asks. Minor gaps: the Problem section does not restate expected behavior in a separate sentence, and the evidence links are external file-drop URLs that reviewers may not be able to access. Neither gap blocks a decision.

Finishing Touches

Generate unit tests (beta)

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Recent review info

Run configuration

• Source mode: Auto
• Repository: pingdotgg/t3code
• Branch: factory/issue-35982
• File: .coderabbit.config.ts
• Commit: 501c45a
• Layers: Repository TS config
• Review profile: CHILL
• Plan: Team
• Run ID: 082caf9f-7f31-43f3-afa0-2959c299d30d

Commits

Reviewing files that changed from the base of the PR and between 43ce395 and 501c45a.

Files selected for processing: 2

• apps/server/src/device/SimlockDeviceHost.test.ts
• apps/server/src/device/SimlockDeviceHost.ts

Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 0 remain after this review.