flutter-testing
Skill16 starsChanged 2 months agoBSD-3-ClauseView on GitHub
Design, write, or review Flutter unit, widget, golden, and integration tests. Use for test strategy, regressions, fakes, async UI behavior, and end-to-end flows; use device testing for operational emulator or hardware control.
7 other projects use a version of this.
Adopted by
Most recently taken up by
- rolandtolnay/flutter-llm-toolkittheir own version, first seen Nov 8, 2026
- growerp/growerptheir own version, first seen Nov 1, 2026
- thienqo/flutter-skillsword for word, first seen Nov 1, 2026
- gspencergoog/skillstheir own version, first seen Oct 16, 2026
- Ethiel97/flutter-clean-riverpod-skillstheir own version, first seen Oct 2, 2026
- gabuldev/flutter-skillstheir own version, first seen Sep 17, 2026
Other versions7 projects use a version of this.
Word for word · 1
- thienqo/flutter-skills★ 16100% same
Edited · 0
None found.
Wrote their own version · 6
- growerp/growerp★ 115same name
- MADTeacher/mad-agents-skills★ 106same name
- gabuldev/flutter-skills★ 12same name
- Ethiel97/flutter-clean-riverpod-skills★ 10same name
- gspencergoog/skills★ 2same name
- rolandtolnay/flutter-llm-toolkit★ 2same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository39 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- dart-concurrencySkill
- dart-languageSkill
- flutter-accessibilitySkill
- flutter-ai-integrationSkill
- flutter-animationSkill
- flutter-app-workflowSkill
- flutter-architectureSkill
- flutter-authenticationSkill
- flutter-background-executionSkill
- flutter-build-releaseSkill
- flutter-ci-cdSkill
- flutter-code-reviewSkill
- flutter-dependency-upgradesSkill
- flutter-device-testingSkill
- flutter-figma-workflowSkill
- flutter-in-app-purchasesSkill
- flutter-localizationSkill
- flutter-mediaSkill
- flutter-navigationSkill
- flutter-networkingSkill
- flutter-notificationsSkill
- flutter-observabilitySkill
- flutter-openapi-clientSkill
- flutter-package-developmentSkill
- flutter-performanceSkill
- flutter-persistenceSkill
- flutter-platform-integrationSkill
- flutter-product-analyticsSkill
- flutter-responsive-layoutSkill
- flutter-runtime-debuggingSkill
- flutter-securitySkill
- flutter-state-managementSkill
- flutter-text-renderingSkill
- flutter-ui-designSkill
- flutter-ui-patternsSkill
- flutter-visual-effectsSkill
- flutter-webviewSkill
The file
We split this skill into the rules it gives an agent
7 rules. 0 are common, 6 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 6 edited versions we read keep these.
6
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Test observable behavior at the cheapest layer that provides confidence.
- Distinctive: line 17
Do not replace behavior assertions with implementation details such as private method calls or exact widget-tree shape.
- Distinctive: line 22
Prefer small fakes over large mock graphs.
- Distinctive: line 26
Update a golden only after reviewing the rendered difference.
- Distinctive: line 34
Prefer an isolated temporary project or directory outside the repository when the runner permits it.
- Very specific: line 35
If the probe must live under test/ or integration_test/, choose one explicit unique path, record whether it existed or was tracked before …
- Distinctive: line 37
Delete only artifacts created by the current run.
“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Flutter Testing
Test observable behavior at the cheapest layer that provides confidence.
Choose the layer
- Unit tests: pure logic, state transitions, repositories, parsing, and failure policy.
- Widget tests: rendering, semantics, interaction, focus, navigation wiring, and state-to-UI behavior.
- Golden tests: stable visual contracts where platform/font control makes comparison meaningful.
- Integration tests: critical flows, plugin integration, performance, and behavior requiring a real target.
Do not replace behavior assertions with implementation details such as private method calls or exact widget-tree shape.
Reliability
- Inject clocks, randomness, storage, and transports when determinism matters.
- Prefer small fakes over large mock graphs; verify outputs and side effects rather than incidental calls.
- Control async progress with the test framework; avoid arbitrary sleeps.
- Pump until a known state or bounded completion, not an unbounded
pumpAndSettlewhen animations or timers persist. - Give tests independent data and clean up global bindings, files, databases, and subscriptions.
- Update a golden only after reviewing the rendered difference.
Disposable test probes
Classify each new test before creating it. Keep regression, acceptance, and specification tests as normal project files. A passing test is not a reason to delete it.
For a one-off diagnostic probe that should not be committed:
- Prefer an isolated temporary project or directory outside the repository when the runner permits it.
- If the probe must live under
test/orintegration_test/, choose one explicit unique path, record whether it existed or was tracked before creation, refuse to overwrite it, and mark the file as generated and disposable. - Register host-side cleanup as soon as the path is created so it runs after success or failure. Preserve the test exit status and useful output before cleanup.
- Delete only artifacts created by the current run. Never recursively delete a test directory, use a broad glob, or delete pre-existing or tracked tests as routine cleanup.
- Confirm the recorded paths no longer exist and compare scoped
git statuswith the pre-run state before reporting completion or preparing a commit. If execution was interrupted, inspect again rather than assuming cleanup ran.
Use test-framework teardown such as addTearDown for runtime files, databases, bindings, and subscriptions created inside a test. It does not replace host-side removal of an ephemeral Dart source file. If a probe exposes a reproducible defect, prefer converting it into a durable regression test unless the user explicitly wants only transient evidence.
Load test layers and doubles when choosing boundaries, fakes, plugin seams, or native-UI coverage. Load determinism, goldens, integration, and cleanup for async flakiness, visual contracts, device flows, or disposable test artifacts.
Verification
Use the Dart and Flutter MCP test capability when available or the repository's established test command otherwise. Run the narrow test during iteration, then the relevant suite. Report the execution surface and data source independently—for example, a widget test with a fixture or an iOS simulator calling staging. Report skipped platforms, flaky behavior, or environment dependencies instead of treating them as success.
Sources
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.

