mochammadshenna/go-adk-q/testing-strategy

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Design test strategies and test plans. Use when deciding how to test a new system, auditing existing test coverage, writing a test plan for a feature, or determining the right balance of unit, integration, and end-to-end tests.

Install

npx vskill@latest install mochammadshenna/go-adk-q/testing-strategy

SKILL.md

Testing Strategy

Design effective testing strategies balancing coverage, speed, and maintenance.

Testing pyramid

/ E2E \ Few, slow, high confidence
/ Integration \ Some, medium speed
/ Unit Tests \ Many, fast, focused

More unit tests than integration tests. More integration tests than E2E. Invert this and your test suite is slow and fragile.

Strategy by component type

ComponentUnit testsIntegration testsE2E / smoke
Business logicYes — table-driven, edge casesNoNo
API endpointsYes — handler logicYes — HTTP layerSmoke only
Database accessNoYes — against real DBNo
CLI / TUIYes — model logicYes — key sequence testsManual QA
Data pipelinesYes — transformation logicYes — idempotency, orderingSmoke only

What to test

Definitely test:

  • Business-critical paths — if it breaks, users notice immediately
  • Error handling — especially error paths that are rarely exercised manually
  • Edge cases — empty input, nil, zero, max values, concurrent access
  • Security boundaries — auth checks, input sanitisation
  • Data integrity — transforms that must be lossless

Skip:

  • Trivial getters and setters
  • Framework code (test your use of the framework, not the framework itself)
  • One-off scripts not used in production
  • Tests that only verify that a mock was called

Test quality criteria

A good test:

  1. Has a clear name that describes what it's testing: TestRenderMessage_AgentRole_RendersWithPromptStyle
  2. Has one logical assertion (it can be multiple t.Error calls, but one concept)
  3. Fails for only one reason
  4. Runs in < 1 second without network or disk I/O
  5. Is deterministic — same result every run

A bad test:

  1. Tests implementation, not behaviour (breaks on refactor)
  2. Is flaky — sometimes passes, sometimes fails
  3. Tests multiple unrelated things
  4. Requires a specific environment to run

For go-adk-q

See cmd/tui/mdtest_test.go for the project's test conventions:

  • Table-driven with t.Run
  • testModel() helper to create a model at a specific terminal size
  • assertContains / assertNotContains helpers for viewport content
  • Race detector always on: go test -race ./...

New tests go in _test.go files alongside the code they test. Use package main (same package) for white-box tests, package main_test (external) for black-box.

Output format

markdown
## Test Plan — [Feature/Component]
### What to test
[List of behaviours to verify]
### Test types
[Unit / integration / E2E for each area, with rationale]
### Coverage targets
Minimum: X% | Target: Y%
### Example test cases
[2-3 concrete examples showing test name and what it verifies]
### Existing coverage gaps
[What's currently untested and why it matters]

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

Repository stars
1
7-day trend score
0

Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/mochammadshenna/go-adk-q/testing-strategy/badge)](https://verified-skill.com/skills/mochammadshenna/go-adk-q/testing-strategy)

Something wrong with this skill or its page? Tell us.