Select Playwright test tiers and capture passed-count evidence for plan close-out.
npx skills add https://github.com/jonaseck2/slaktforskning --skill e2e-evidence
Please help me install this Agent Skill. Skill: e2e-evidence Source: https://github.com/jonaseck2/slaktforskning/tree/main/.claude/skills/e2e-evidence Command: npx skills add https://github.com/jonaseck2/slaktforskning --skill e2e-evidence
This Skill prevents weak or misleading end-to-end closure by telling you exactly which Playwright tier to run and what test output counts as real evidence before archiving.
[boot] / [crud] / [panels] / [reactivity] / [imports] fail or are flaky, route the fix to the owning plan or require a separate plan explicitly covering the fix before archive.Decide which tier fits your change and paste the required “npm run test:e2e” and, if needed, “npm run test:e2e:full” tail output lines into your plan close-out evidence.
High-intent search queries and answers about installing and using this skill.
Playwright tier selection for plan close-out depends on which user goals your changes touch. Tier 1 always runs across boot, crud, website-export, and duplicates, while Tier 2 adds panels, reactivity, and imports when those specific UI surfaces are modified.
Valid e2e test evidence for CI release gating requires the actual passed-count tail lines from your test command runs, such as the specific "N passed" output, rather than subjective summaries or claims of success in your close-out notes.
Triage CI failures by Playwright project tags like boot, crud, panels, reactivity, or imports by routing the fix to the owning plan before archive, or by requiring a separate plan explicitly covering the fix.
If your changes only touch basic crud operations, running test:e2e alone suffices because Tier 1 deterministically covers boot, crud, website-export, and duplicates. You only need test:e2e:full for Tier 2 when UI, panels, data-changed, or importer user goals are touched.
You should not archive a plan after running Playwright e2e tests if the close-out lacks the actual tail line passed-counts from the test commands, or if project-tagged failures remain flaky without a separate fix plan.
Map a new e2e spec to the correct Playwright project by matching the user-observable surface it protects—such as boot, crud, panels, reactivity, or imports—to the project-to-user-goal mapping before adding the spec to that tier.
cloudflareStandardize unit and E2E testing conventions for Sandbox SDK projects.
mattjeffersonVerify release readiness by checking test evidence, invariants, and rollback feasibility.
ksciusRun Playwright smoke tests and diagnose failures with traces and logs.
lespacemanGenerate a prioritized e2e-plan/coverage-plan.md with P0/P1/P2 coverage and TC-IDs.