agentleFS
LedgerHQLedgerHQ/ledger-live/.agents/skills/e2e-desktop-onboard/SKILL.md

e2e-desktop-onboard

Skill627 starsChanged 10 days agoMITView on GitHub

Interactive setup wizard for running desktop E2E (Playwright + Speculos) tests locally. Use when a developer sets up their machine to run desktop E2E tests for the first time.

  • Pipes a download into a shell
  • Reads credentials
  • Deletes or force-pushes
  • Installs packages

The file

Show its rules and text

We split this skill into the rules it gives an agent

22 rules. About 0 are common, 17 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈17

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    CRITICAL SECURITY RULE: NEVER print, log, or display the value of SEED or any secret.

    line 17
  2. Distinctive:

    Do NOT hardcode expected versions: Read the source-of-truth config files to determine what's required:

    line 25
  3. Distinctive:

    PASS: if the active version matches the pinned version exactly.

    line 46
  4. Very specific:

    WARN: if the active major version matches but the minor/patch differs (e.g.

    line 47
  5. Distinctive:

    FAIL: only if the active major version is different or older than the pinned major version (e.g.

    line 48
  6. Very specific:

    Check each variable is set and valid.

    line 64
  7. Very specific:

    If COINAPPS is set and the directory exists, check if the repo is up to date:

    line 85
  8. Distinctive:

    If the fetch shows new commits, warn the user and suggest:

    line 91
  9. Distinctive:

    Verify the Docker + Speculos setup is ready.

    line 105
  10. Distinctive:

    If it fails, Docker Desktop is not running.

    line 111
  11. Distinctive:

    If the pull fails:

    line 117
  12. Distinctive:

    Check the image tag is correct

    line 120
  13. Distinctive:

    If behind a proxy, Docker may need proxy configuration

    line 121
  14. Distinctive:

    If any containers are listed, they are leftover from previous test runs (likely interrupted with Ctrl+C).

    line 137
  15. Distinctive:

    Run sequentially (check exit codes between each step).

    line 151
  16. Distinctive:

    If any step fails, report the error and help the user fix it before continuing.

    line 183
  17. Distinctive:

    Run a quick test to validate the full Speculos setup:

    line 196
  18. Distinctive:

    Do not treat it as a failed setup step.

    line 213
  19. Distinctive:

    If the app launches and connects to Speculos successfully but the test fails on a specific UI element (e.g.

    line 236
  20. Very specific:

    Use ⚠️ manual for the smoke test row when the GUI fallback was triggered (Electron could not launch from the agent shell).

    line 248
  21. Distinctive:

    If everything passed, confirm the user is ready to run tests with:

    line 265
  22. Distinctive:

    If anything failed, list the remaining issues and offer to help fix them.

    line 289

≈ “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

Desktop E2E Onboarding

You are an interactive setup wizard that checks the developer's machine and guides them through every prerequisite for running Ledger Live Desktop E2E tests (Playwright + Speculos) locally.

This command sets up Speculos mode (MOCK=0) exclusively -- real hardware wallet simulation via Docker. Mock mode does not require onboarding.

Work through each phase sequentially. For each check, run the shell command, report pass/fail, and if it fails, guide the user through the fix before moving on. Use TodoWrite to track progress across phases.

All commands assume the working directory is the repository root (ledger-live/). Before Phase 1, verify with git rev-parse --show-toplevel and cd there if needed.

CRITICAL SECURITY RULE: NEVER print, log, or display the value of SEED or any secret. Only confirm whether it is set or not.

DO NOT MODIFY TEST CODE: This command is strictly for environment setup and validation. Never create, edit, or delete any test file (specs, page objects, helpers, fixtures, or any file under e2e/).


Phase 1: System Tools

Do NOT hardcode expected versions. Read the source-of-truth config files to determine what's required:

ToolSource of truthHow to read
Nodemise.tomlgrep "^node" mise.toml
pnpmmise.tomlgrep "^pnpm" mise.toml

First, read the config files to determine expected versions. Then run these checks:

CheckCommandExpected
Dockerdocker --versionInstalled (required -- Speculos runs in Docker)
Docker runningdocker info > /dev/null 2>&1Docker Desktop is running
Misemise --versionInstalled (required)
Nodenode --versionSee version policy below
pnpmpnpm --versionSee version policy below

Node / pnpm version policy:

The [tools] section in mise.toml is the source of truth for the recommended versions. The check should:

  • PASS if the active version matches the pinned version exactly.
  • WARN if the active major version matches but the minor/patch differs (e.g. pinned 22.13.1, active 22.14.0). This is acceptable and must not block the setup or progression to the next phase.
  • FAIL only if the active major version is different or older than the pinned major version (e.g. pinned Node 22, active Node 20). Only FAIL results are hard blockers that must be resolved before proceeding.

The version is what matters, not how Node or pnpm were installed. If versions drift, suggest running mise install from the repo root (after installing mise).

Fix commands for common failures:

Do not proceed to Phase 2 until all relevant FAIL-level checks are resolved. WARN-only results (Node/pnpm minor/patch drift versus mise.toml) should be surfaced but must not block progression.


Phase 2: Environment Variables

Check each variable is set and valid. Run [ -n "$VAR" ] && echo "set" || echo "NOT SET" for each.

VariableValidationDefault / Guidance
SEEDNon-emptyAsk the team for guidance
MOCKEquals 0 ([ "$MOCK" = "0" ])Must be 0 for Speculos mode
COINAPPSNon-empty AND directory exists ([ -d "$COINAPPS" ])Path to cloned coin-apps repo. If not cloned: git clone https://github.com/LedgerHQ/coin-apps.git
SPECULOS_IMAGE_TAGNon-emptyghcr.io/ledgerhq/speculos:latest
SPECULOS_DEVICENon-emptynanoSP (options: nanoS, nanoSP, nanoX, stax, flex, nanoGen5)

Optional variables:

VariablePurposeDefault
SPECULOS_FIRMWARE_VERSIONFirmware version to emulateAuto-detected from device model
DISABLE_TRANSACTION_BROADCASTSet to 1 to prevent real tx broadcastUnset (broadcast enabled)

For any missing variable (except SEED), provide the exact export line to add to ~/.zshrc.

coin-apps freshness check:

If COINAPPS is set and the directory exists, check if the repo is up to date:

cd "$COINAPPS" && git fetch origin master --dry-run 2>&1

If the fetch shows new commits, warn the user and suggest:

cd "$COINAPPS" && git pull origin master

Outdated coin-apps is a common cause of Speculos app version mismatches.

Do not proceed to Phase 3 until all relevant checks pass.


Phase 3: Docker and Speculos

Verify the Docker + Speculos setup is ready.

3a. Docker Desktop running

Run: docker info > /dev/null 2>&1

If it fails, Docker Desktop is not running. Guide the user to open it and wait.

3b. Pull Speculos image

Run: docker pull $SPECULOS_IMAGE_TAG

If the pull fails:

  • Check internet connectivity
  • Check the image tag is correct
  • If behind a proxy, Docker may need proxy configuration

3c. Verify Speculos image works

Run: docker run --rm $SPECULOS_IMAGE_TAG --help

This should print the Speculos help text. If it fails, the image may be corrupted -- suggest removing and re-pulling:

docker rmi $SPECULOS_IMAGE_TAG && docker pull $SPECULOS_IMAGE_TAG

3d. Check for stale Speculos containers

Run: docker ps -a --filter name=speculos --format "{{.ID}} {{.Names}} {{.Status}}"

If any containers are listed, they are leftover from previous test runs (likely interrupted with Ctrl+C). Offer to clean them up:

docker rm -f $(docker ps -a --filter name=speculos -q)

Stale containers can cause port conflicts when launching new Speculos instances.

Do not proceed to Phase 4 until all checks pass.


Phase 4: Install Dependencies and Build

Run sequentially (check exit codes between each step). All commands from the repo root.

4a. Install dependencies

pnpm i --filter="ledger-live-desktop..." --filter="live-cli..." --filter="ledger-live" --filter="@ledgerhq/dummy-*-app..." --filter="ledger-live-desktop-e2e-tests" --unsafe-perm

4b. Build lib dependencies

pnpm build:lld:deps

This also builds the CLI, a desktop dev dependency.

4c. Build desktop app for testing

pnpm desktop build:testing

This creates a production-like Electron build with TESTING=1 (uses .env.testing Firebase config).

4d. Install Playwright browser

pnpm e2e:desktop test:playwright:setup

This installs Chromium for Playwright.

If any step fails, report the error and help the user fix it before continuing.

Common build failures:

  • OOM: prefix command with NODE_OPTIONS="--max-old-space-size=10240"
  • Stale build artifacts: rm -rf apps/ledger-live-desktop/.webpack then retry pnpm desktop build:testing
  • Chromium download fails: check network/proxy, retry pnpm e2e:desktop test:playwright:setup
  • Dependency conflicts after branch switch: pnpm clean then start from step 4a again

Phase 5: Smoke Test

Run a quick test to validate the full Speculos setup:

pnpm e2e:desktop test:playwright tests/specs/add.account.spec.ts

Report whether the test passed or failed. If it failed, check the error output and determine which category the failure falls into.

GUI environment fallback:

Electron is a GUI application that requires macOS WindowServer access. The agent shell may not have this access. If the test output contains any of these patterns:

  • Process failed to launch!
  • Cannot read properties of undefined (reading 'getAppPath')

This is a GUI environment limitation, not a setup failure. In this case:

  1. Do not treat it as a failed setup step.
  2. Report that all setup phases (1-4) completed successfully and the environment is ready.
  3. Explain that Electron cannot launch from the agent shell due to missing display/WindowServer access.
  4. Print the exact command for the user to run in their Cursor integrated terminal (not the agent shell):
Your E2E environment is fully set up. To validate, paste this command in your Cursor integrated terminal:

  pnpm e2e:desktop test:playwright tests/specs/add.account.spec.ts
  1. Mark the smoke test as manual in the Phase 6 summary (not as a failure).

Common smoke test failures (real setup issues):

  • "Timeout waiting for selector": app didn't load -- check that pnpm desktop build:testing ran successfully, check that apps/ledger-live-desktop/.webpack/main.bundle.js exists
  • Docker errors / Speculos container failed to start: check Docker is running (docker info), check COINAPPS path is correct, check image tag matches a pulled image
  • TypeError: Invalid Version: DS_Store: .DS_Store files in the coin-apps repo are causing semver lookup failures. Fix with: find $COINAPPS -name ".DS_Store" -type f -delete
  • Speculos app version mismatch / missing nano app: the coin-apps repo may be outdated. Pull latest: cd $COINAPPS && git pull origin master
  • Port conflicts: stale Speculos containers from previous runs. Clean up with: docker rm -f $(docker ps -a --filter name=speculos -q)

Setup vs flow mismatch:

If the app launches and connects to Speculos successfully but the test fails on a specific UI element (e.g. a button not found after an unexpected screen appears), the problem is likely a test-flow mismatch, not a broken setup. In this case:

  • Confirm the setup is valid (app launched, Speculos connected, Docker container running).
  • Report that the failure appears to be a test scenario issue, not an environment issue.
  • Suggest the user check the test expectations against the current app flow.

Phase 6: Summary

Print a final checklist summarizing everything.

Use ⚠️ manual for the smoke test row when the GUI fallback was triggered (Electron could not launch from the agent shell). This indicates the setup is complete but the smoke test must be confirmed by the user in their terminal.

## E2E Desktop Setup -- Complete

| Check                  | Status              |
| ---------------------- | ------------------- |
| Docker                 | ✅ / ❌             |
| Mise / Node / pnpm    | ✅ / ❌             |
| Environment variables  | ✅ / ❌             |
| Speculos image         | ✅ / ❌             |
| Dependencies installed | ✅ / ❌             |
| App built              | ✅ / ❌             |
| Playwright browser     | ✅ / ❌             |
| Smoke test             | ✅ / ❌ / ⚠️ manual |

If everything passed, confirm the user is ready to run tests with:

You're all set! Run tests from the repo root:

  # All tests
  pnpm e2e:desktop test:playwright

  # Single test file
  pnpm e2e:desktop test:playwright tests/specs/<testFile>.spec.ts

  # Pattern filter
  pnpm e2e:desktop test:playwright --grep "swap"

  # Debugging (Playwright inspector + Ledger Live devtools)
  PWDEBUG=1 DEV_TOOLS=1 pnpm e2e:desktop test:playwright tests/specs/<testFile>.spec.ts

  # Disable transaction broadcast
  DISABLE_TRANSACTION_BROADCAST=1 pnpm e2e:desktop test:playwright tests/specs/<testFile>.spec.ts

  # Generate and view Allure report after a test run
  pnpm e2e:desktop allure

If anything failed, list the remaining issues and offer to help fix them.

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.