Where the minutes go (30d)
Waste detected
Biggest wins first, each with the exact config fix.
Workflow failing almost every run · tests 2
Fails or times out on nearly all recent runs, burning minutes to produce only red, and an always-red workflow usually gets ignored. Fix it, or disable the trigger until it's ready.
Missing dependency cache · tests 2
Set `cache: 'gradle'` on your `actions/setup-java` step, or add an `actions/cache@v4` step keyed on your build file.
- uses: actions/setup-java@v4
with:
distribution: 'temurin'
java-version: '21'
cache: 'gradle'Missing dependency cache · tests 1
Set `cache: 'pnpm'` on your `actions/setup-node` step, or add an `actions/cache@v4` step keyed on your lockfile.
- uses: actions/setup-node@v4
with:
node-version: 20
cache: 'pnpm'No concurrency control · tests 2
Add a `concurrency:` block keyed on branch to cancel superseded runs when devs push twice quickly.
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
cancel-in-progress: trueMissing dependency cache · tests BiDi
Set `cache: 'npm'` on your `actions/setup-node` step, or add an `actions/cache@v4` step keyed on your lockfile.
- uses: actions/setup-node@v4
with:
node-version: 20
cache: 'npm'Missing dependency cache · infra
Set `cache: 'pip'` on your `actions/setup-python` step, or add an `actions/cache@v4` step keyed on your requirements/lock file.
- uses: actions/setup-python@v5
with:
python-version: '3.12'
cache: 'pip'Missing dependency cache · Publish Test Results
Set `cache: 'npm'` on your `actions/setup-node` step, or add an `actions/cache@v4` step keyed on your lockfile.
- uses: actions/setup-node@v4
with:
node-version: 20
cache: 'npm'Missing dependency cache · Update test results DB
Set `cache: 'npm'` on your `actions/setup-node` step, or add an `actions/cache@v4` step keyed on your lockfile.
- uses: actions/setup-node@v4
with:
node-version: 20
cache: 'npm'No concurrency control · infra
Add a `concurrency:` block keyed on branch to cancel superseded runs when devs push twice quickly.
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
cancel-in-progress: trueNo concurrency control · Publish Test Results
Add a `concurrency:` block keyed on branch to cancel superseded runs when devs push twice quickly.
concurrency:
group: ${{ github.workflow }}
# no cancel-in-progress on deploys: queued deploys would be droppedMissing dependency cache · extension
Set `cache: 'npm'` on your `actions/setup-node` step, or add an `actions/cache@v4` step keyed on your lockfile.
- uses: actions/setup-node@v4
with:
node-version: 20
cache: 'npm'Also found: 24 more config fixes with negligible recoverable minutes: timeouts, retention, path filters
No job timeout · infra
No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.
jobs:
build:
runs-on: ubuntu-latest
timeout-minutes: 15No path filters on triggers · infra
Runs on every push/PR with no `paths:` filter, so docs-only changes still trigger full CI. Add a `paths:` filter if that's common.
on:
pull_request:
paths:
- 'src/**'
- 'package.json'No job timeout · publish release - trace viewer
No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.
jobs:
build:
runs-on: ubuntu-latest
timeout-minutes: 15Premium runners (macOS / Windows) · tests 1
macOS bills ~10x and Windows ~2x a Linux minute. The cost estimate above assumes Linux, so your real spend is higher. Move any job that doesn't need them to `ubuntu-latest`.
jobs:
build:
runs-on: ubuntu-latest # ~10x cheaper than macos-latestenvironment: on a CI job · tests 1
A test/build job is bound to a GitHub environment, so every run counts as a deployment: PR and deployments-API noise. Use repository secrets or OIDC for non-deploy jobs.
No path filters on triggers · tests 2
Runs on every push/PR with no `paths:` filter, so docs-only changes still trigger full CI. Add a `paths:` filter if that's common.
on:
pull_request:
paths:
- 'src/**'
- 'package.json'Premium runners (macOS / Windows) · tests 2
macOS bills ~10x and Windows ~2x a Linux minute. The cost estimate above assumes Linux, so your real spend is higher. Move any job that doesn't need them to `ubuntu-latest`.
jobs:
build:
runs-on: ubuntu-latest # ~10x cheaper than macos-latestenvironment: on a CI job · tests 2
A test/build job is bound to a GitHub environment, so every run counts as a deployment: PR and deployments-API noise. Use repository secrets or OIDC for non-deploy jobs.
Larger hosted runner · tests 2
Uses a larger hosted runner, which bills even on public repos (only standard runners are free). Keep big runners for jobs that measurably need the cores.
jobs:
lint:
runs-on: ubuntu-latest # standard runner — free on public reposNo job timeout · Publish Test Results
No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.
jobs:
build:
runs-on: ubuntu-latest
timeout-minutes: 15No job timeout · tests BiDi
No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.
jobs:
build:
runs-on: ubuntu-latest
timeout-minutes: 15Scheduled at the top of the hour · tests BiDi
The schedule fires at minute :00, GitHub's peak window, where scheduled runs get delayed or skipped. Shift to any other minute for the same cadence with less contention.
on:
schedule:
- cron: '50 0 * * *' # was '0 0 * * *' — any non-:00 minute avoids the herdScheduled workflow runs in forks · tests BiDi
The schedule has no repository guard, so forks that enable Actions inherit the cron and burn their own minutes. Gate the job with `if: github.repository == 'owner/repo'`.
jobs:
nightly:
if: github.repository == 'owner/repo'
runs-on: ubuntu-latestSystem packages reinstalled every run · tests BiDi
System packages install from the network on every run with no cache. Cache them (cache-apt-pkgs-action) or check whether the runner image already has the tool.
- uses: awalsh128/cache-apt-pkgs-action@latest
with:
packages: <your packages>
version: 1.0Tests wrapped in auto-retry · tests BiDi
Tests run under an auto-retry wrapper, so each retry re-bills the same minutes and masks the flaky tests it papers over. Track, fix or quarantine the flakes, then drop the wrapper.
environment: on a CI job · tests BiDi
A test/build job is bound to a GitHub environment, so every run counts as a deployment: PR and deployments-API noise. Use repository secrets or OIDC for non-deploy jobs.
No job timeout · MCP
No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.
jobs:
build:
runs-on: ubuntu-latest
timeout-minutes: 15Premium runners (macOS / Windows) · MCP
macOS bills ~10x and Windows ~2x a Linux minute. The cost estimate above assumes Linux, so your real spend is higher. Move any job that doesn't need them to `ubuntu-latest`.
jobs:
build:
runs-on: ubuntu-latest # ~10x cheaper than macos-latestenvironment: on a CI job · MCP
A test/build job is bound to a GitHub environment, so every run counts as a deployment: PR and deployments-API noise. Use repository secrets or OIDC for non-deploy jobs.
No job timeout · extension
No job sets `timeout-minutes`, so a hung step can run to GitHub's 6-hour default. Add `timeout-minutes` to each job.
jobs:
build:
runs-on: ubuntu-latest
timeout-minutes: 15Premium runners (macOS / Windows) · extension
macOS bills ~10x and Windows ~2x a Linux minute. The cost estimate above assumes Linux, so your real spend is higher. Move any job that doesn't need them to `ubuntu-latest`.
jobs:
build:
runs-on: ubuntu-latest # ~10x cheaper than macos-latestMissing dependency cache · Issue triage
Set `cache: 'npm'` on your `actions/setup-node` step, or add an `actions/cache@v4` step keyed on your lockfile.
- uses: actions/setup-node@v4
with:
node-version: 20
cache: 'npm'Artifacts at default retention · Issue triage
`upload-artifact` has no `retention-days`, so artifacts keep up to 90 days (storage cost). Set e.g. `retention-days: 7`.
- uses: actions/upload-artifact@v4
with:
name: build
path: dist/
retention-days: 7Scheduled at the top of the hour · Update test results DB
The schedule fires at minute :00, GitHub's peak window, where scheduled runs get delayed or skipped. Shift to any other minute for the same cadence with less contention.
on:
schedule:
- cron: '6 */3 * * *' # was '0 */3 * * *' — any non-:00 minute avoids the herdCopy every fix as one block (15 snippets, annotated per workflow)
# Missing dependency cache (gradle) (applies to: tests 2)
- uses: actions/setup-java@v4
with:
distribution: 'temurin'
java-version: '21'
cache: 'gradle'
# Missing dependency cache (pnpm) (applies to: tests 1)
- uses: actions/setup-node@v4
with:
node-version: 20
cache: 'pnpm'
# No concurrency control (applies to: tests 2, infra)
concurrency:
group: ${{ github.workflow }}-${{ github.ref }}
cancel-in-progress: true
# Missing dependency cache (npm) (applies to: tests BiDi, Publish Test Results, Update test results DB, extension, Issue triage)
- uses: actions/setup-node@v4
with:
node-version: 20
cache: 'npm'
# Missing dependency cache (pip) (applies to: infra)
- uses: actions/setup-python@v5
with:
python-version: '3.12'
cache: 'pip'
# No concurrency control (applies to: Publish Test Results)
concurrency:
group: ${{ github.workflow }}
# no cancel-in-progress on deploys: queued deploys would be dropped
# No job timeout (applies to: infra, publish release - trace viewer, Publish Test Results, tests BiDi, MCP, extension)
jobs:
build:
runs-on: ubuntu-latest
timeout-minutes: 15
# No path filters on triggers (applies to: infra, tests 2)
on:
pull_request:
paths:
- 'src/**'
- 'package.json'
# Premium runners (macOS / Windows) (applies to: tests 1, tests 2, MCP, extension)
jobs:
build:
runs-on: ubuntu-latest # ~10x cheaper than macos-latest
# Larger hosted runner (applies to: tests 2)
jobs:
lint:
runs-on: ubuntu-latest # standard runner — free on public repos
# Scheduled at the top of the hour (applies to: tests BiDi)
on:
schedule:
- cron: '50 0 * * *' # was '0 0 * * *' — any non-:00 minute avoids the herd
# Scheduled workflow runs in forks (applies to: tests BiDi)
jobs:
nightly:
if: github.repository == 'owner/repo'
runs-on: ubuntu-latest
# System packages reinstalled every run (applies to: tests BiDi)
- uses: awalsh128/cache-apt-pkgs-action@latest
with:
packages: <your packages>
version: 1.0
# Artifacts at default retention (applies to: Issue triage)
- uses: actions/upload-artifact@v4
with:
name: build
path: dist/
retention-days: 7
# Scheduled at the top of the hour (applies to: Update test results DB)
on:
schedule:
- cron: '6 */3 * * *' # was '0 */3 * * *' — any non-:00 minute avoids the herdEach snippet is representative. Merge into the named workflow files rather than pasting wholesale.
Live CI-health badge → 
[](https://gitspider.com/scan/microsoft/playwright)Want this on every push?
This scorecard is a one-time snapshot. Install the free GitHub App to track this repo continuously: new regressions caught as they land, trends over time, on your public and private repos. Team adds the offending commit on the PR + Slack alerts.
Install & monitor this repo →💬 On Team, this same analysis posts automatically to every PR: the regression, and the exact commit that caused it, right where your team already looks. See plans →
Not ready to install? Get this report by email. No spam, unsubscribe anytime.