test(browser): bound ordinary teardown to one deadline

[seonghobae]

Partial implementation of #28, stacked on live PR #147 branch test/agent-task-forced-close-shared-teardown-deadline.

Falsifiable defect

Ordinary controlled Agent Task teardown previously waited for the exact browser-root identity and the already sampled Chromium process set with two independent bounded waits. A slow or stuck teardown could therefore consume the configured timeout twice even though the product evidence describes one bounded teardown phase.

RED -> RCA -> GREEN

1. Test-first contracts require ordinary teardown to share one monotonic deadline across exact root-identity termination and sampled process-set termination while preserving those as separate evidence fields.
2. The pre-fix contract exposed that the two helpers each created or consumed their own full timeout budget; sequential invocation could extend nominal teardown latency without changing the caller-visible timeout.
3. The smallest production remedy computes one ordinary-teardown deadline after session/driver shutdown and passes only the remaining budget to each observation boundary. The existing fail-closed root/process-set evidence semantics remain separate, and None remains preserved when no complete process set was captured.
4. CHANGELOG.md records this exact behavioral boundary and its limits; no cgroup/task ownership, post-snapshot process discovery, cross-platform supervision, or OS-wide orphan-absence claim is added.

Fresh stack reconstruction

The live prerequisite #147 moved to exact 962e179b53ed0342fde5b50c3deac0e7454b0d4f after this child was last validated. The child was therefore reconstructed non-destructively from GitHub's current conflict-free synthetic merge tree, preserving both the prerequisite changes and this PR's unique ordinary-teardown delta.

Current exact head: f42b388d92c03a2f2ccbe7ca07803e5f5d2aea44.
Exact prerequisite/base head: 962e179b53ed0342fde5b50c3deac0e7454b0d4f.
Fresh ancestry comparison: 7 commits ahead / 0 behind, with merge base exactly the current prerequisite head.

The previous exact-head CI/coverage/MV3 evidence belongs to predecessor 4da223ac612ea52c04022f5380919ce85a868d15 and is not promoted to the reconstructed head. Fresh exact-head checks and review evidence must complete on f42b388d92c03a2f2ccbe7ca07803e5f5d2aea44 before this child can advance.

Truth boundary

This proves only bounded ordinary controlled Linux/pinned-Chromium teardown observation for the exact browser root and the already sampled Chromium process set. It does not claim cgroup/task ownership, discover processes created after the sample, prove OS-wide orphan absence, authenticate Chromium/ChromeDriver provenance, implement cross-platform supervision, or turn this CI compatibility runner into a product browser API.

The scheduled actor does not merge to protected main, self-approve, force-push, alter workflows/rulesets/secrets, tag, release, or publish.

---

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 0f8b55ae-a146-486b-a90e-9160da0048a3

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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