test-design
Skill0 starsChanged 5 days agoMITView on GitHub
Use this when test cases must be designed from a spec, detail design or user story (already converted with doc-to-md) — choosing test techniques per requirement, building a traceable test case list with expected results that a tester or Playwright run can execute, and showing coverage gaps.
23 other projects use a version of this.
Adopted by
Most recently taken up by
- LeanMeanCleanApps/CleanLambdaFunctiontheir own version, first seen Nov 15, 2026
- LeanMeanCleanApps/CleanWebApitheir own version, first seen Nov 15, 2026
- LeanMeanCleanApps/CleanGraphQLApitheir own version, first seen Nov 15, 2026
- LeanMeanCleanApps/CleanMinimalApitheir own version, first seen Nov 15, 2026
- selamy-labs/agent-skillstheir own version, first seen Nov 8, 2026
- Agenta-AI/agentatheir own version, first seen Nov 8, 2026
Used with
Projects with it also use diff-review.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
MidnightDarlingSkilldiff-review
使用场景:按 raw.review.md 清单改完稿后,对 raw.md(OCR 原始)和 final.md(定稿)做对比。说"看我改了哪些""对比一下""diff""校对完了让我看看""我到底接受了哪些建议""我漏改了啥""改动统计""核对一下改动""闭环检查"都走这个 skill。生成段落级 HTML diff,关联 raw.review.md 的 A/B/C 标注,一目了然:接受了哪些 agent 建议、拒绝了哪些、清单外修正了哪些、agent 标过但漏改了哪些。主动触发,不必等用户说 "diff-review"。
Also in 3 projects that use this
Other versions23 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 23
- Agenta-AI/agenta★ 4.9ksame name
- LeanMeanCleanApps/CleanMinimalApi★ 400same name
- adamentwistle/fable-skills★ 98same name
- nahid-sparktales/agent-dispatcher★ 59same name
- LeanMeanCleanApps/CleanGraphQLApi★ 13same name
- shabaraba/vibing.nvim★ 13same name
- tlzmw001/aitest-kit★ 13same name
- yasunori0418/skills★ 9same name
- umararshad4/honey-stack★ 5same name
- LeanMeanCleanApps/CleanWebApi★ 4same name
13 more not listed
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 repository24 other files it gives its agents.
- anatomySkill
- approachSkill
- brainstorm-brseSkill
- business-explainSkill
- change-impactSkill
- code-to-businessSkill
- cross-checkSkill
- detail-design-jpSkill
- diagramSkill
- doc-to-mdSkill
- e2e-test-playwrightSkill
- estimationSkill
- excel-analyzeSkill
- image-analyzeSkill
- jp-vi-translateSkill
- meeting-intakeSkill
- proposal-genSkill
- qa-tone-brseSkill
- report-genSkill
- research-jp-techSkill
- review-gateSkill
- scope-analysisSkill
- testcase-review-jpSkill
- verify-outputSkill
The file
We split this skill into the rules it gives an agent
4 rules. 0 are common, 1 is distinctive, and 3 are very specific to this project.
0
Common
Most of the 17 edited versions we read keep these.
1
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 14
Run <py> ${CLAUDE_PLUGIN_ROOT}/scripts/brse.py config and read its JSON:
- Very specific: line 26
After the documents are in Markdown (doc-to-md) and before execution (e2e-test-playwright).
- Very specific: line 52
Always add: required-field empty, permission/role negatives, error message wording, back/refresh/double-click, and data persistence after …
- Distinctive: line 72
Test case file path, counts (cases, per priority, automatable), the coverage gaps and the Q&A list inline.
“Common” is measured: kept by at least 6 in 10 of the 17 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
Before running
<py> below = the Python recorded by /brse-toolkit:setup (runtime.python
in ~/.config/brse-toolkit/config.json — read that file directly), else the
first of python3, python, py -3 that runs. No Python at all → tell the
user to run /brse-toolkit:setup, and continue only with steps that need no
script, saying which checks were skipped.
Run <py> ${CLAUDE_PLUGIN_ROOT}/scripts/brse.py config and read its JSON:
error→ show it to the user and stop; the config file must be fixed first.workspace_rootis null → the folder isn't initialized. Ask once: run/brse-toolkit:initfirst (recommended — memory and knowledge get saved for next time) or continue once without saving. Running as a subagent dispatched by a main agent → the parent already resolved this; proceed.- Otherwise use
config.paths(sources,templates,outputs— relative toworkspace_root),config.style,config.translation, and the files in.brse/memory/long/. Never assume a folder name the config doesn't list. warningsnot empty → mention them in one line (likely a typo in the config).
When to use
After the documents are in Markdown (doc-to-md) and before execution
(e2e-test-playwright). For reviewing an existing test case list against a
design, use testcase-review-jp instead.
Steps
- Restate scope: feature / screens, test level (unit / integration /
system / E2E / regression), reader, output language and file format
(client template in
config.paths.templatesfirst; else the layout below). - Requirement list with IDs. From
.brse/knowledge/docs/extract every testable statement: input fields with types/limits, business rules, messages, state changes, permissions, screen transitions, non-functional limits. Give each an ID (REQ-<screen>-<n>) and cite the source file + heading/sheet/row. Anything ambiguous → a Q&A item (要確認), not a guessed expected result. Pick the technique per requirement — say why:
Requirement shape Technique Input with ranges / lengths / formats Equivalence partitioning + boundary values (min-1, min, max, max+1, empty, type-wrong) Several conditions → different outcomes Decision table (one case per meaningful column) Object with statuses (order, account lock, approval) State transition: every valid transition + key invalid ones Many independent parameters (browser × role × option) Pairwise, not full combination User journey across screens Use-case / scenario: main flow + each alternative/exception flow Concurrency, 排他制御, double submit Explicit two-session cases Messages, labels, layout Checklist against the spec text, exact wording Always add: required-field empty, permission/role negatives, error message wording, back/refresh/double-click, and data persistence after reload.
- Write the cases. One row = one verifiable result. Columns:
No | REQ ID | 観点/Viewpoint | Precondition | Test data | Steps | Expected result | Priority (H/M/L) | Technique | Automatable (Y/N) | Result | Remark. Steps are concrete actions on named UI elements (soe2e-test-playwrightcan drive them); expected results quote the spec wording, never "works correctly". Test data are real values, including the boundary values. - Traceability + coverage. A matrix REQ ID → case Nos; list every requirement with zero cases and every case with no REQ (remove or justify). Count cases per technique and per priority.
- Self-review with the
revieweragent role: duplicates, untestable expected results, missing negatives. Fix, then hand off. - File: Excel via the toolkit's xlsx scripts (
excel-analyze/report-gen) intoconfig.paths.outputs; never overwrite a client file. Human-readable text for the client (観点, expected-result wording in Japanese/Vietnamese) goes through the translator role (Comtor) per the project's translation config.
Output
Test case file path, counts (cases, per priority, automatable), the coverage gaps and the Q&A list inline.
Handoff
testcase-review-jp (optional second check) → e2e-test-playwright for
execution → report-gen. Finish with review-gate.
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.

