Przejdź do treści
Biblioteka skilli
Programowanie · Przegląd i audyt

test-hygiene

/test-hygiene

Pilnuje, gdzie powstają pliki testów i czy dodać test do istniejącego pliku, oraz porządkuje zestaw przez kwarantannę niestabilnych testów i scalanie powtórzeń.

Licencja
MIT
Ostatnia zmiana repozytorium
27.10.2026

Ocena i bezpieczeństwo

100/ 100Klasa: dobry
Format25 / 25
Frontmatter z nazwą i opisem, poprawna nazwa, opis do 1024 znaków, nazwa zgodna z folderem.
Odkrywalność25 / 25
Opis mówi, kiedy użyć skilla, ma sensowną długość i jest w 3. osobie.
Treść35 / 35
Co najmniej 150 słów i do 500 linii, nagłówki, listy kroków, przykłady albo kod.
Higiena15 / 15
Bez zaślepek typu TODO, bez ścieżek Windows, z licencją.
Ryzykobrak sygnałów w skanie
Pliki w paczce
8
Skrypty
0
Komentarze HTML
0
Adresy URL w SKILL.md
1
Przestarzałość

Brak sygnałów. Repozytorium zmienione 27.10.2026, bez odwołań do wycofanych modeli i API.

Licencja

MIT, otwarta: można używać i zmieniać, z zachowaniem informacji o autorze.

Co robi i kiedy

Pilnuje, gdzie powstają pliki testów i czy dodać test do istniejącego pliku, oraz porządkuje zestaw przez kwarantannę niestabilnych testów i scalanie powtórzeń.

Oryginalny opis (EN)

Test placement rules and the measure, quarantine, consolidate remediation ladder for degraded test suites. TRIGGER WHEN: creating or placing test files, deciding whether to extend an existing test or add a new file, auditing test-suite health, quarantining flaky tests, or consolidating redundant tests. DO NOT TRIGGER WHEN: TDD workflow mechanics (use mattpocock-skills:tdd) or browser E2E page objects/waiting (use developer-essentials:e2e-testing-patterns).

Jak działa

  1. Why suites degrade under agentic coding
  2. The binding rules
  3. The remediation ladder
  4. Layer model and budgets
  5. Related knowledge
Sekcje z nagłówków SKILL.md, w oryginale · 153 linii, 1 797 słów
Podgląd SKILL.md
---
name: test-hygiene
description: >
  Test placement rules and the measure, quarantine, consolidate remediation ladder for degraded test suites.
  TRIGGER WHEN: creating or placing test files, deciding whether to extend an existing test or add a new file, auditing test-suite health, quarantining flaky tests, or consolidating redundant tests.
  DO NOT TRIGGER WHEN: TDD workflow mechanics (use mattpocock-skills:tdd) or browser E2E page objects/waiting (use developer-essentials:e2e-testing-patterns).
---

<daodan_host_binding host="copilot">
This binding governs execution of this package and its loaded references. It takes
precedence over executable examples written for another host. Use only registered
tools and their actual input schema; an example never creates a tool, argument,
permission, model alias or nested worker capability. Preserve the generated worker
allowlist, dependency closure, scope and delivery contracts.

Copilot binding: repository reads use search, writes use edit, shell execution uses
runCommands and network fetch uses fetch in the adapter's tool allowlist. Custom
agent dispatch uses the agent allowlist alias; the current CLI documents a task
dispatch tool. Call the mechanism actually registered in this session, not an
allowlist alias assumed to be a callable tool.
Use each registered tool's input fields; Claude's subagent_type, role and
run_in_background fields are not a portable dispatch schema. Dispatch only the
custom agent named by the generated owner-qualified binding and agents allowlist.
An omitted tools restriction retains host defaults; an explicit list constrains
the worker and does not grant omitted shell, write or MCP tools.
Skills use their exact installed registry identity; workflows are prompt entries.
A dispatched workflow also has its generated <workflow>-coordinator custom agent.
Required foreign providers use Copilot's marketplace/install commands:
copilot plugin marketplace add <source>, then copilot plugin install <provider>@<marketplace>.
Register plugin-declared MCP server commands in this host's MCP configuration as
the generated note directs. Registration and a dependency declaration alone do
not demonstrate that browser/server tools are callable by the selected worker.
Inherit the active session's model and default variant. A source model alias such
as sonnet is not a portable override; request a different model only through an
explicit adapter mapping to a valid host identity and user authorization.
Coordinator MCP allowances use <declared-server>/* for only the package's named
servers. This requires actual server registration and tool discovery; it grants
no undeclared server and does not prove MCP is callable in an installed run.
Schema reference: https://docs.github.com/en/copilot/reference/custom-agents-configuration

The executing agent and its coordinator own continuation, recovery, verification
and delivery. Skills supply knowledge and method instructions; they are not
executors or a runtime persistence mechanism.

Resolve each component by kind and exact installed identity. A role is dispatched
as a worker, a skill loads instructions, and a workflow starts its declared method.
Kernel owner-qualified skill references are logical provider identities: resolve
them to the actual host registry name and verified installed provider path. Do not
invent a registry string by adding an owner's prefix when the host does not use it.
Do not load an agent identity as a skill or execute a workflow name as a shell
command. Declared required foreign providers remain mandatory. If a required
provider or capability (including required MCP access or isolation) is unavailable,
stop its dependent phase and report the missing capability. Optional external
providers follow the method's declared selection and absence behavior; their
absence alone does not block the base workflow. Report any explicitly requested
but unavailable provider contribution as unfulfilled and continue independent
work. Do not install a different host's plugin or
substitute unrelated skills to claim coverage.

Shell execution alone does not establish interactive stdin. Before a capture or
checkpoint, use an available session-control mechanism. When stdin is unavailable,
keep user-driven capture in a supported background process with bounded output and
an explicit stop signal controlled by the coordinator after the user's completion
message. Close owned browser/process resources on completion or cancellation.

Continue working within the current authorization while the requested result is
achievable. After a tool error, inspect its cause, correct the relevant input or
implementation, retry when justified and verify the outcome; never repeat a
failing action blindly. Do not close the turn with a promise or a partial result
when an authorized next step can still produce or verify the requested result.
Stop only for a documented real impediment, a missing required capability, an
unresolved human decision or the caller's stop request. At that boundary, report
completed results, remaining work and the concrete prerequisite for continuation;
continue any independent work that the blocker does not prevent.

A tool's successful exit, a worker stop or an idle session is not the deliverable.
Return final text naming observed findings, artifact paths, executed checks and
blockers. An empty worker result is missing delivery: inspect or resume that worker
when permitted, or record failure. Verify required artifacts and candidate gates
before marking the requested outcome complete.

This is an execution binding, not a rewrite of source examples. Preserve programming
language fences, domain code and host APIs that are the subject of marketplace-ops
or agent-sdk-builder. A host-as-subject API is not a tool provided to this session.
Compiled binding and static checks do not prove an installed host executed a run.
</daodan_host_binding>

# Test-Suite Hygiene

Rules and workflows that keep a test suite small, trustworthy, and navigable while agents and humans keep adding to it.

## Why suites degrade under agentic coding

For an agent, writing a new test file costs nearly nothing; understanding the existing suite costs context. The dominant strategy becomes "create a new one", and the implicit goal is "task closed, CI green", never "suite healthy". Nobody measures suite health, so nobody optimizes it. The result compounds fast: parallel test files for the same module, the same behavior asserted at three layers, implementation-coupled tests that break on every refactor and get replaced instead of repaired, and skip markers that quietly turn the safety net into decoration. These rules exist to make the healthy move the cheap move.

## The binding rules

For new or changed behavior, identify the introduced failure modes and the smallest
meaningful checks that protect them. Reuse sufficient existing protection; extend
it where needed. Trivial changes need no ceremonial test. Report unverified
behavior and remaining risk.

Test ownership: source at unit; behavior at integration, contract and e2e.

Full protocol with per-rule detail in `references/prevention-rules.md`. The condensed form:

1. **Review before adding.** Locate and read the existing owner for the intended layer and behavioral scope. Map protection before adding any case: reuse, update or justify the gap. Zero additions is valid. A parallel file requires a justified scope split.
2. **Deterministic ownership per layer.** Unit tests: one primary test file per source file, mirroring the source path (`src/foo/bar.py` maps to `tests/unit/foo/test_bar.py`), following the project's established convention and documenting justified scope splits. Integration, contract, and e2e tests are behavior-owned: one primary file per behavioral scope (a flow, an endpoint, a contract), legitimately spanning several source modules. The violation at those layers is an unexplained second file for the same scope, not multi-module reach.
3. **Explicit layers** (unit, integration, contract, e2e) with runtime budgets, identified through the project's directories, colocated paths or runner markers. A new test goes in the lowest layer that can prove its contract without hiding the relevant boundary.
4. **Behavior through justified interfaces.** Prefer observable contracts. Internal access, boundary doubles and ordering checks require a named boundary or diagnostic regression; inspect their purpose before rewriting them.
5. **No skip markers to get green.** Classify product, oracle, environment and intermittent failures before repair or accepted temporary quarantine. A tracked reason alone does not justify losing regression protection.

Assertion protection: Preserve assertions for approved behavior; correct a wrong oracle only with independent authority and evidence, while keeping valid product protection active.

Test retirement: Retire tests only for retired behavior or verified equivalent replacement protection in the same candidate; counts, coverage, age and refactoring alone are insufficient.

Refactors, renames and migrations preserve and retarget meaningful protection.
Resource-dependent tests follow the isolation and reproduction checklist in
`references/prevention-rules.md`, rule 9.

## The remediation ladder

When the suite has already degraded, rules alone do not repair it. Bonify in layers, opportunistically, alongside normal development. Full mechanics in `references/remediation-workflow.md`.

1. **Measure first.** The audit mode of `testing:test-remediation-method` produces a versioned `TEST_AUDIT.md`: counts, runtime, skipped, failing, flaky, orphans, layer distribution, slowest tests, per-module coverage. Metrics guide investigation; they do not establish behavioral value or correctness.
2. **Classify before quarantine.** Load the `testing:test-remediation-method` skill: product defects, wrong oracles, environment failures, intermittence and unknown causes have different owners. Temporary quarantine has explicit evidence, remaining risk, return condition and owner; a green suite alone does not become trustworthy.
3. **Consolidate per module.** Inventory protected behavior, independent failure modes and bugfix provenance before rewrite. Accepted replacements and removal share one candidate and authorized commit.
4. **Verify.** Every protected inventory row maps to a justified surviving check. Coverage and counts support the assessment but cannot prove equivalence.

Order remediation by risk (production bugs, then churn from git log), not by how messy a module looks.

## Layer model and budgets

Defaults; a project's own convention overrides them.

| Layer | Contains | Budget (default) |
|---|---|---|
| `unit` | Isolated behavior, controlled doubles at justified boundaries, no real external I/O | Individual test under 100ms; whole layer under 60s |
| `integration` | Real boundaries (DB, HTTP, filesystem) via containers or fixtures | Whole layer under 5min |
| `e2e` | Critical user flows only | Project-defined runtime budget; justified critical flows and failure modes, no test quota |

A behavior's primary proof lives at ONE layer. Cross-layer overlap is duplication only when two tests protect substantially the same failure mode through the same observable contract without adding independent risk coverage; a calculation checked at unit, its persistence at integration, and the user flow at e2e are three behaviors, not one repeated. A test that crosses a real database boundary belongs in integration. A controlled database-port double can isolate service behavior at unit, but cannot prove SQL, transactions or persistence; those contracts require the real boundary in an appropriate test lane.

## Related knowledge

Source revisions, license notices and delegation boundaries are indexed in
`references/upstream.md`; read it for provenance or integration maintenance.

Required upstream skills need a capability preflight on the active host: resolve
and load the actual installed skill before accepting an action that needs it.
Dependency metadata and five-host rendering do not establish that an external
marketplace runs on all five hosts. If unavailable, record the affected action
open and give the current host's supported installation guidance; do not substitute
another prompt or claim the upstream method ran. Claude-specific install examples
below apply only on that host.

- Writing the test content (red-to-green workflow, behavior-first design, mocking discipline): load the `mattpocock-skills:tdd` skill. It ships in the upstream mattpocock/skills plugin, a hard dependency of this plugin; if it is unavailable, stop and tell the user to install it (`claude plugin marketplace add mattpocock/skills`, then `claude plugin install mattpocock-skills@mattpocock`).
- Browser E2E patterns (page objects, fixtures, waiting, network mocking): load the `developer-essentials:e2e-testing-patterns` skill. It ships in the upstream wshobson/agents marketplace, a hard dependency of this plugin; if it is unavailable, stop and tell the user to install it (`claude plugin marketplace add wshobson/agents`, then `claude plugin install developer-essentials@claude-code-workflows`).
- Runner detection and measurement commands for the audit workflows: `references/runner-playbook.md`.
Pliki w paczce (8)
  • SKILL.md 13 KB
  • SKILL.toml 1 KB
  • references/LICENSE-mattpocock-skills-MIT.txt 1 KB
  • references/LICENSE-wshobson-agents-MIT.txt 1 KB
  • references/prevention-rules.md 16 KB
  • references/remediation-workflow.md 8 KB
  • references/runner-playbook.md 10 KB
  • references/upstream.md 9 KB

Wersje

RepozytoriumWynikLicencja
Ta strona · acaprino/alfio-claude-plugins100MIT
acaprino/alfio-claude-plugins100MIT
acaprino/alfio-claude-plugins100MIT
acaprino/alfio-claude-plugins100MIT
acaprino/alfio-claude-plugins100MIT
acaprino/alfio-claude-plugins100MIT
checkpickerupper/skills100MIT

Identyczna treść występuje jeszcze w 1 innych folderach na GitHubie. Tu pokazujemy jedną, najlepiej ocenioną kopię.

Migawka bazy z 13.11.2026 · wersja paczki 58cef5458c14 (skrót SHA-256 całej paczki)

Pobrania w ostatnich 30 dniach: 1 · ranking

Zgłoś problem z tym skillem: szkodliwa treść, błędny opis albo prośba autora o usunięcie.

Podobne skille

  • 85proof
    Działa jako inżynier QA i testów: strategia testów, zestawy E2E, testy API, triage niestabilnych testów i pokrycie.
  • 100testing
    Opisuje strategie testów, w tym TDD i testy E2E, i pomaga w pisaniu testów oraz konfiguracji infrastruktury testów E2E.
  • 100test-strategy
    Podejmuje decyzje na poziomie zestawu testów: skąd brać dane (fabryki czy anonimizowane zrzuty), polityka kwarantanny testów niestabilnych i granica integracji z E2E.
  • 100testing-strategy
    Opisuje metodykę testów: piramidę, deterministyczne testy jednostkowe i integracyjne oraz diagnozę niestabilnych testów.
  • 90test-flakiness-triage
    Przegląda raporty niestabilnych testów, historię powtórzeń CI i testy zawodzące sporadycznie, przypisuje każdemu kategorię przyczyny źródłowej oraz ścieżkę naprawy lub kwarantanny, wyłącznie statycznie.
  • 100testing
    Opisuje, jak organizować testy, pokrycie, izolację, testy regresyjne i TDD, oraz wymagania dotyczące wykonania testów.

Uwagi i zgłoszenia

Wiadomość trafi do Bartosza Sokolińskiego. Adres e-mail jest potrzebny tylko do odpowiedzi. Szczegóły w polityce prywatności.