[whi-tw]

What problem does this pull request solve?

Trello card: https://trello.com/c/jPeqHHPu/772-improve-the-observability-of-our-github-workflows

If a workflow fails on main or in the merge queue, nobody is told unless they happen to look. This adds a reusable workflow that posts to #govuk-forms-alert when that happens. Failures on pull requests are not reported, because the author already sees them. notify-failure.yml lists the workflows it watches by name, so a new workflow that runs on main or in the merge queue needs adding to that list.

It also sets timeout-minutes on every job, so a hung job fails fast rather than running for up to six hours, and adds a check to the shared workflow lint that fails if a job has none. forms-deploy was not running that lint on itself, so it now calls it.

The limits are deliberately generous (10 to 30 minutes) against p95 durations of roughly two minutes or less.

Before merging

• An org-level SLACK_WEBHOOK_GOVUK_FORMS_ALERT secret exists and is available to the forms-* repos.
• Run "Notify failure" manually (workflow_dispatch) to check a test message arrives.
• Open the equivalent PRs in forms-admin, forms-runner, forms-product-page and forms-e2e-tests straight after merging. They run this repo's lint at @main, so their workflow changes will fail it until they have timeouts.

Things to consider when reviewing

• Do the timeouts look sensible for each job?
• Is the workflow_run approach, with a list of watched workflows, acceptable compared with adding a notify job to each workflow?

[Copilot]

🟡 Changes recommended

The notification watch list omits the new workflow lint, leaving its merge-queue failures unreported.

1 open issue

• Add Lint Workflows to the Slack notification watch list · New

What changed in this PR

Improves GitHub Actions observability with Slack failure alerts and explicit job timeouts.

Changes:

• Adds reusable Slack notifications for watched workflows failing on main or in the merge queue.
• Sets job timeouts across existing workflows.
• Enforces timeouts through shared workflow linting and enables that lint locally.

File	Description
.github/​workflows/​update-provider-locks.yml	Adds a 10-minute timeout.
.github/​workflows/​trello-poster.yml	Adds a 10-minute timeout.
.github/​workflows/​terraform-ci.yml	Adds 20-minute job timeouts.
.github/​workflows/​team-review-requests.yml	Adds a 10-minute timeout.
.github/​workflows/​shell-ci.yml	Adds a 10-minute timeout.
.github/​workflows/​ruby-ci.yml	Adds 15-minute job timeouts.
.github/​workflows/​reusable-review_apps_on_pr_close.yml	Limits cleanup to 20 minutes.
.github/​workflows/​reusable-review_apps_on_pr_change.yml	Limits updates to 30 minutes.
.github/​workflows/​reusable-notify-failure.yml	Adds reusable Slack message delivery.
.github/​workflows/​reusable-lint-workflows.yml	Enforces job timeouts and adds its own.
.github/​workflows/​reusable-build-image.yml	Limits builds to 30 minutes.
.github/​workflows/​require-pinned-github-actions.yml	Adds a 10-minute timeout.
.github/​workflows/​pre-commit.yml	Adds 10-minute job timeouts.
.github/​workflows/​pipeline-visualiser.yml	Adds 20-minute job timeouts.
.github/​workflows/​notify-failure.yml	Watches failures and supports test notifications.
.github/​workflows/​lint-workflows.yml	Runs shared lint for pull requests and merge groups.
.github/​workflows/​guardrail_unique_job_names.yml	Adds a 10-minute timeout.
.github/​workflows/​guardrail_tf_version_upgrade.yml	Adds a 10-minute timeout.
.github/​workflows/​guardrail_matching_pentester_email_addresses.yml	Adds a 10-minute timeout.
.github/​workflows/​guardrail_matching_pentester_cidr_blocks.yml	Adds a 10-minute timeout.
.github/​workflows/​forms-cli.yml	Adds a 15-minute timeout.

_{Review effort: Balanced}

[Copilot]

🔵 Needs a closer look

Slack delivery, secret availability, and coordinated rollout to repositories consuming the shared lint require human confirmation.

No open issues 🎉

1 resolved since last review

• Add Lint Workflows to the Slack notification watch list

_{Review effort: Balanced}

[Copilot]

🔵 Needs a closer look

Slack delivery and the coordinated rollout of stricter shared linting to consuming repositories require human verification.

No open issues 🎉

_{Review effort: Balanced}

[Copilot]

🔵 Needs a closer look

Slack delivery needs manual verification, and shared timeout enforcement requires a coordinated rollout across consuming repositories.

No open issues 🎉

_{Review effort: Balanced}

[github-actions]

Warning

You are changing the Terraform version.

Before you merge this PR, you must apply the forms/pipelines
Terraform root from this branch in each of the forms environments.

make ENV forms/pipelines apply

[Copilot]

🟡 Changes recommended

The notification secret references must match the provisioning checklist for Slack delivery to work under the documented setup.

1 open issue

• Align Slack webhook secret name with pre-merge checklist · New

_{Review effort: Balanced}

[Copilot]

🔵 Needs a closer look

Slack delivery, secret availability, and the coordinated rollout to shared-lint consumers still require human verification.

1 open issue

• Align Slack webhook secret name with pre-merge checklist

_{Review effort: Balanced}