ng-verify-feature
Skill13 starsChanged 44 days agoMITView on GitHub
Verify a running Angular feature against acceptance criteria using the browser, and report observed behavior with reproduction evidence. Use when the user asks to verify a delivered feature, check an Angular user flow, or perform scoped exploratory acceptance testing. Do not use for source-only code review, writing e2e specs, or a general explanation of browser testing.
In this repository34 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .junie/AGENTS.mdAGENTS.md
- .claude/CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- angular-developerSkill
- brainstormingSkill
- code-reviewSkill
- create-a-skillSkill
- create-e2e-testsSkill
- diagnosing-bugsSkill
- frontend-designSkill
- git-stack-rewriteSkill
- grill-meSkill
- grill-with-styleSkill
- handoverSkill
- implement-planSkill
- ng-accessibilitySkill
- ng-data-accessSkill
- ng-formsSkill
- ng-migrateSkill
- ng-performanceSkill
- ng-prototypeSkill
- ng-refactorSkill
- ng-review-architectureSkill
- ng-review-style-guideSkill
- ng-securitySkill
- ng-signal-storeSkill
- ng-stylingSkill
- ng-testingSkill
- spartanSkill
- test-driven-developmentSkill
- unslopSkill
- update-skills-directorySkill
- update-skillsSkill
The file
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 8 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈8
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
- Distinctive: line 17
Read the task, approved design, or relevant acceptance criteria and the project's browser rules.
- Distinctive: line 23
Use pass, fail, blocked, or not tested.
- Distinctive: line 30
Check the project's configured URL for an existing server.
- Distinctive: line 34
Do not start a dev server or Chrome, install a browser runtime, or switch to an internal browser without the required user authorization.
- Distinctive: line 39
Use existing authenticated state where appropriate.
- Distinctive: line 51
Check the relevant cases from the coverage table:
- Distinctive: line 60
Use available fixtures or authorized request interception to reach error states.
- Distinctive: line 101
Check those three outcomes in Chrome, capture failures at the point of interaction, and report any unreachable state as blocked.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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.
Files
Verify an Angular Feature
Adapted from Vercel's dogfood: checks a bounded feature against acceptance criteria using available
Chrome tools, replaces issue quotas with coverage, and preserves the user's browser and server sessions.
Upstream license: references/LICENSE.txt.
1. Define what must work
Read the task, approved design, or relevant acceptance criteria and the project's browser rules. Identify the target route, changed behavior, and relevant user state. If criteria are implicit, derive a small list of observable outcomes and state those assumptions. Ask only when an unresolved choice would materially change what counts as correct.
Prepare a compact coverage table: criterion, action, expected result, observed result, status, evidence. Use pass, fail, blocked, or not tested; empty evidence is never a pass. Keep exploration within the feature and its directly affected flows.
Done when the target and observable acceptance criteria are explicit.
2. Connect to the existing app
Check the project's configured URL for an existing server. Use the user-managed Chrome session and available Chrome DevTools or browser tools, following the tool's connection instructions. Do not assume agent-browser, Playwright, Cypress, or a particular debugging connection is installed or available.
Do not start a dev server or Chrome, install a browser runtime, or switch to an internal browser without the required user authorization. When a prerequisite is missing, report exactly what is unavailable and request it. You may inspect source and run permitted static checks while waiting, but label them separately from browser evidence. Never treat a missing browser as a successful verification.
Use existing authenticated state where appropriate. Keep test mutations within the authorized scope and disposable test data; do not send messages, make purchases, or change real user data merely to exercise a flow. Do not export the user's authentication state into report files.
Done when the intended route is reachable in the permitted browser, or dependent checks are blocked.
3. Exercise the feature
Inspect the rendered page before acting. Locate controls by accessible role, label, or observed text; refresh the page state after navigation or substantial UI changes. Wait for the specific expected UI state, not a fixed sleep or a universal network-idle condition.
Check the relevant cases from the coverage table:
- The main user journey and its persisted result, where persistence is part of the requirement.
- Empty, loading, error, disabled, and populated states that the feature actually supports.
- Form validation, keyboard operation, visible focus, and accessible labels.
- Navigation, direct links, and back/forward behavior when routes or query parameters changed.
- Relevant viewport sizes and realistic content lengths when layout changed.
- Console errors and failed requests associated with the exercised action.
Use available fixtures or authorized request interception to reach error states; record when a case cannot be exercised. An automated accessibility scan supplements keyboard and visual checks. Do not claim full WCAG conformance from a scan or from this bounded feature check.
Done when every planned case has an observed result or a stated reason it could not be tested.
4. Capture findings as they happen
Record expected versus actual behavior, the route and relevant state, and minimal reproduction steps while the context is fresh. Attach a screenshot, DOM/accessibility observation, or console/network excerpt that demonstrates the result. Record video only when timing or interaction makes it useful. Exclude credentials and sensitive content from evidence.
Retry an apparent failure when safe. Distinguish reproducible failures from intermittent observations; retain uncertain observations with that qualification. Classify severity by user impact, and distinguish acceptance failures from optional UX suggestions. Zero findings is a valid result; do not invent issues to meet a quota.
Done when each finding has evidence and enough context for another person to investigate it.
5. Report coverage and limitations
Return the coverage table and severity-ranked findings with evidence links. State the tested route, user state, browser, and material limitations. Use the project's report location if one exists; create a small report file only when the evidence or number of cases makes it useful.
A verification request does not authorize code changes. When fixes are already authorized, repair within that scope and rerun affected cases before updating their status. Use the project's e2e skill only when reusable test authoring is requested; browser inspection does not require adding a test runner.
Restore temporary viewport or test-state changes you made when safe, and leave the user's existing browser and server running. Clean up only disposable data created for this verification.
Done when the user can see which criteria passed, failed, were blocked, or remain untested, with no completion claim stronger than the evidence.
Example
“Verify the booking filter on the running app: cancelled bookings disappear, clearing the filter restores them, and the filter is usable with the keyboard.”
Check those three outcomes in Chrome, capture failures at the point of interaction, and report any unreachable state as blocked. A passing unit test alone does not satisfy these browser criteria.
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
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.

