agentleFS
Sign inStart free
robinbrilrobinbril/claude-code-harness/skills/e2e-rules/SKILL.md

e2e-rules

Skill3 starsChanged 23 days agoMITView on GitHub

House rules for end-to-end UI tests with e2e (TesterArmy). Use alongside the e2e skill on any UI change in a web or mobile app, when setting up e2e in a project, and when an e2e run fails. Covers the subscription-only model setup, the config template, file naming, and when a task counts as done.

  • Installs packages

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

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

    Test files are named *.e2e.ts and live in tests/.

    line 10
  2. Very specific:

    After changing a test, run it once with --no-cache.

    line 14
  3. Very specific:

    Never put secrets in tests.

    line 15

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

e2e house rules

  • On a UI change in a web or mobile app: add or update an e2e test. Use the e2e skill and the e2e MCP server.
  • Project without e2e: copy assets/e2e.config.template.ts from this skill to e2e.config.ts, adjust targets (start command, URL), then npm i -D e2e@0.21.2 @e2e-dev/web@0.15.2 ai@7.0.143. The template runs models on your own Claude seat through claudeSeat() from poort, which ships with this harness: node <harness>/tools/poort/poort.mjs init (v0.4.0); fix the relative import of .poort/integrations/e2e-claude-seat.mjs if the config sits in a subfolder. Without poort, set the model in the config as the e2e skill describes. Do not run npx e2e init over an existing config.
  • Test files are named *.e2e.ts and live in tests/.
  • Models run on the user's own Claude seat: claudeSeat() from .poort/integrations/e2e-claude-seat.mjs runs claude -p. No API keys, no proxy, no other providers. Switch models with E2E_MODEL / E2E_JUDGE (Claude ids, e.g. claude-sonnet-5-5, claude-opus-5-5). Not signed in: ask the user to run claude and log in, or to set CLAUDE_CODE_OAUTH_TOKEN from claude setup-token. Recording sees demo data only; replay calls no model.
  • One agent.act per goal, followed by an exact expect.
  • Done = npx --no-install e2e run --reporter list,markdown passes. On failure, read .e2e/summary.md and the failure page before changing code.
  • After changing a test, run it once with --no-cache. Commit .e2e/cache/.
  • Never put secrets in tests; use credentials.user().

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.

Check your setup, free