Command

PePersonal JEJust Electrolyte RTRelease Timer TtTacTile CSClearSource AtAthletic

TacTile›Launch TacTile›DevOps›DevOps Foundation·35 in this wave

Approve: DevOps FoundationTo do · blocker NEEDS MATTHEW: staging STRIPE_WEBHOOK_SIGNING_SECRETBlocked · blocker main is RED — 8 unit tests failing on origin/main, so pnpm verify cannot gate any PRDone · blocker The production account's data is 42 days stale — Analytics renders empty at any default windowBlocked · blocker MATTHEW: add db_url to the 1Password Supabase item — two RLS audits have never been runnableDone · blocker Wire apps/api test suite into CIDone Wire Playwright specs into CIDone Wire verify:swift-parity into CIDone Enable branch protection on mainBlocked Verify the staging environment is actually liveDone Web deploy workflow (manual trigger only)To do API deploy workflow (manual trigger only)To do Add post-deploy smoke test to the deploy workflowsTo do Write the rollback procedureTo do Add Sentry to iOSTo do Add Sentry to AndroidTo do Configure uptime monitoringTo do Set the three missing staging secrets on tactile-api-stagingDone Playwright is failing on main and producing no report — every run that gets a runner failsTo do Playwright job must free port 3000 BEFORE its first suite — a watchdog-killed job leaks its server and fails every later runDone Add CHANGELOG.mdTo do Automate release tagging on mainTo do Fix stale Vercel references in the observability runbookDone Add apps/api typecheck to CIDone Clean the SAST baselineTo do Flip SAST to blockingTo do Edge function deploy workflow (manual trigger only)To do Supabase migration check in CITo do Write the release procedureTo do Configure Supabase log drainTo do Centralize Cloudflare Workers logsTo do Replace console.* with structured logging in apps/apiTo do Two runners share one ~/.gradle — scope it per runner without losing the warm cacheTo do MATTHEW: CI capacity — one Mac, two runners, and a full Xcode build now runs on every pushTo do Auto-cancel queued CI runs whose head branch was deletedTo do

Wave

DevOps Foundation

Unblocked now — nothing waits on the surfaces. Wiring existing tests into CI, deploy automation, base observability.

TacTile › Launch TacTile › DevOps › DevOps Foundation

10/35 done25 open3 open blockers

Selected taskclear

MATTHEW: CI capacity — one Mac, two runners, and a full Xcode build now runs on every push

To doOwner · met_361002e9

No dependencies in this wave — nothing gates it, and nothing waits on it.

Show the prompt

You are Viv, the agent for the TacTile lane of Command.

Open the lane and work there:

```bash
cd ~/Command/tactile && claude
```

Read that lane's CLAUDE.md and lane.json before you touch anything — they
carry your identity, your scope, and the hazards specific to this lane.
Read state from the hub (https://internal.tactile.to) or ask. Never guess at it.

## The task

**MATTHEW: CI capacity — one Mac, two runners, and a full Xcode build now runs on every push** · to do

Where it sits: TacTile › Launch TacTile › DevOps › DevOps Foundation

## The brief on file

HARD NUMBERS ADDED 2026-08-16. This was filed as a judgement call ("CI feels tight"); it now has a measurement, and the measurement is worse than the intuition.

MEASURED over the 40 most recent `verify` runs since #485 wired the Playwright job in:
  40 playwright jobs created
   6 ever assigned a runner  (15%)
  34 cancelled with `runner_name` EMPTY — they never started
   6 of 6 that executed PASSED

So 85% of the time that job reports a red X for a verdict it never measured. Measured
instance: job 95195407046 queued 16:46:09, cancelled 16:49:16, never assigned.

MECHANISM — three facts multiplying, no single file at fault:
  · the job's concurrency group is the CONSTANT `playwright-e2e-supabase`, which is
    deliberate: the Supabase stack and port 3000 are host-wide, and a ref-scoped group
    is precisely what cost #461 62 spurious failures via cross-run teardown;
  · GitHub holds at most ONE pending job per group and cancels the PREVIOUSLY PENDING
    one when a third arrives;
  · two self-hosted runners stay busy far longer than the gap between pushes during a
    merge wave — a `verify` gate plus an iOS Debug+Release build occupy both for tens
    of minutes.

The pending job is superseded before a runner frees. Nothing inside the job can fix
it; it loses a race it never enters.

PARTIAL MITIGATION SHIPPED (#494): a nightly 09:00 UTC / 02:00 PT schedule on
verify.yml, so the suite executes at least daily when both runners are idle. That is a
guarantee about CONTENTION, not capacity — it does NOT raise the 15% on PRs, and if
the Mac is asleep at 09:00 UTC the run queues and GitHub expires it after 24h.

THE ACTUAL DECISION, which is Matthew's:
  (a) a third self-hosted runner — REJECTED ONCE ALREADY on evidence: three runners
      put load at 235 on 10 cores and produced phantom `verify:unit` /
      `ts-source-coverage` / `app-typecheck` failures that three sessions reported as
      findings. None were real. More runners on ONE Mac makes contention worse, not
      better.
  (b) a second machine — real capacity, real money.
  (c) GitHub-hosted minutes for the cheap Linux jobs only (verify, semgrep,
      static-firewall), keeping macOS work self-hosted. Costs money but is the
      smallest spend that actually unblocks the queue.
  (d) accept 15% and rely on the nightly.

RELATED: t_e7bfca45 (two runners share one ~/.gradle) is the same class of problem —
one machine pretending to be several.

## How to work this

- Do the work, don't just plan it. If the fix is obvious and safe, make it.
- Surface anything needing Matthew's sign-off instead of deciding it yourself.
- Short replies. Lead with the answer, then stop.
- Finish by saying what changed, what you verified, and what you'd do next.

To doIn progressDoneBlockedPick a wave on the left · tap a task to expand it

Lanes

DevOps Foundation

MATTHEW: CI capacity — one Mac, two runners, and a full Xcode build now runs on every push