agentleFS
LedgerHQLedgerHQ/ledger-live/.agents/skills/ci-runners/SKILL.md

ci-runners

Skill627 starsChanged 10 days agoMITView on GitHub

Pick a self-hosted runner for a GitHub Actions job. Read before adding or changing any `runs-on` (or a runner label passed as an input or matrix value) in `.github/workflows/**`.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not add new usages of these labels.

    line 10
  2. Distinctive:

    Never pick silently. Ask the user to choose from the allowed list, with a recommendation:

    line 44

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

CI Runners

🚫 Deprecated runners

Do not add new usages of these labels. When touching a job that already uses one, offer to migrate it.

  • ledger-live-medium
  • ledger-live-4xlarge
  • ledger-live-linux-8CPU-32RAM
  • ledger-live-linux-16CPU-64RAM
  • ledger-live-linux-e2e-8CPU-32RAM
  • ledger-live-linux-e2e-speculos-32CPU-128RAM

Exception: Android emulator jobs. The emulator needs /dev/kvm, which the runners below don't have (see Gotchas). Android E2E jobs in test-mobile-e2e-reusable.yml run on ledger-wallet-8cpu-github-public-cached, a GitHub larger runner on the ledger-wallet-android-e2e custom image (built in LedgerHQ/ledger-wallet-runners), which holds the toolchain and the E2E AVD. Use it only for jobs that run the emulator.

Exception: Windows desktop E2E. The Windows leg of e2e-tests-extra-os in test-ui-e2e-only-desktop.yml runs on ledger-live-windows-e2e-32CPU-128RAM, a Windows runner Platform Engineering provisioned for LWD E2E. Its egress IP is allow-listed on Speculinho, which other GitHub-hosted runners can't reach. It is not deprecated despite the ledger-live- prefix. Use it only for Windows desktop E2E.

✅ Allowed runners

SizeCached (default)Uncached (release)
1 CPUledger-wallet-1cpu-aws-public-cachedledger-wallet-1cpu-aws-public
2 CPUledger-wallet-2cpu-aws-public-cachedledger-wallet-2cpu-aws-public
4 CPUledger-wallet-4cpu-aws-public-cachedledger-wallet-4cpu-aws-public
8 CPUledger-wallet-8cpu-aws-public-cachedledger-wallet-8cpu-aws-public
16 CPUledger-wallet-16cpu-aws-public-cachedledger-wallet-16cpu-aws-public

For jobs that need a Docker daemon, use the 4 CPU Docker runner: ledger-wallet-4cpu-aws-public-docker-cached by default, or ledger-wallet-4cpu-aws-public-docker for the release process.

Memory is 4 GiB per CPU (e.g. 4cpu = 16 GiB, 8cpu = 32 GiB).

Cached vs uncached

  • Default to the -cached variant.
  • Use the uncached variant for the release process: release-*.yml workflows and jobs that run on main, release or hotfix. Release builds must start from a clean state.

Picking a runner

Never pick silently. Ask the user to choose from the allowed list, with a recommendation:

  1. Cached or uncached, from the rule above.
  2. Size from the job's needs. When replacing a deprecated runner, start from the same CPU count (e.g. ledger-live-linux-8CPU-32RAM → 8 CPU). There is no 32 CPU option: propose 16 CPU and flag the job for benchmarking.
  3. -docker only if the job runs containers.

Gotchas

  • Pods report the host CPU count (nproc = 32), not their quota. Never size workers as a percentage of cores (e.g. JEST_MAX_WORKERS: "95%"); set explicit counts per runner size.
  • These runners have no /dev/kvm, so the Android emulator cannot run on them.
  • Forks cannot use self-hosted runners. Keep the existing fork fallback pattern: runs-on: ${{ github.event.pull_request.head.repo.fork && 'ubuntu-22.04' || 'ledger-wallet-4cpu-aws-public-cached' }}
  • Reusable workflows are called at @develop, so a PR cannot prove a runner change. Validate with workflow_dispatch from the branch.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.