e2e_verify: step-9 cancel wait uses 30s default, too tight for a real checker

[mithro]

Summary

The e2e_verify harness's step 9 ("Cancelling precheck...") waits for the
check to reach the Cancelled state using the default 30s timeout. Against
a real checker VM, cancelling a running precheck reliably takes longer than 30s,
so step 9 fails with:

TimeoutError: Timed out waiting for cancellation (30000 ms)

This blocks the harness from ever reaching the manufacturability verdict
(steps 10–13) unless the timeout is bumped locally.

Root cause

scripts/e2e_verify/main.py:299 calls wait_for_precheck_cancelled() with no
timeout, so it uses the short default:

log(9, "Cancelling precheck...")
detail_page.click_cancel_precheck()
detail_page.wait_for_precheck_cancelled()   # <- defaults to 30_000 ms

wait_for_precheck_cancelled (scripts/e2e_verify/pages/project_detail.py:304)
defaults to 30s:

def wait_for_precheck_cancelled(self, timeout_ms: int | None = None) -> None:
    """Wait for precheck cancellation to be confirmed."""
    timeout = timeout_ms or 30_000
    ...

The preflight cancel path already uses 180s and works fine —
cancel_check_if_running (same file, line 369) calls
wait_for_precheck_cancelled(timeout_ms=180_000). Only the step-9 call site
uses the short default.

Suggested fix

Pass a 180s timeout at the step-9 call site to match the working preflight path:

detail_page.wait_for_precheck_cancelled(timeout_ms=180_000)

(Alternatively, raise the default in wait_for_precheck_cancelled itself so
both call sites get the longer window.)

Impact / context

Test tooling only — no production impact. Surfaced repeatedly while running
quarter-tile (0p5x0p5) e2e_verify runs against both stage and prod on
2026-07-30; a local 180s bump let the run complete to a manufacturable verdict.

[coderabbitai]

🔗 Related PRs

#123 - Increase precheck timeout to 12 hours for stage/prod [merged]
#132 - feat: Add E2E verification script with Playwright and VNC [merged]
#182 - config: increase precheck timeout from 12 to 20 hours [merged]
#186 - fix: set Docker client timeout to 5 minutes [closed]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!