test-design
Skill1 starsChanged 4 days agoMITView on GitHub
Use for TDD and choosing, keeping, merging or deleting tests: acceptance, property, contract, golden and end-to-end. Not for behaviour-preserving refactors or root cause; use refactor or diagnose.
18 other projects use a version of this.
Adopted by
Most recently taken up by
- selamy-labs/agent-skillstheir own version, first seen Nov 8, 2026
- Agenta-AI/agentatheir own version, first seen Nov 8, 2026
- LeanMeanCleanApps/CleanModelContextProtocoltheir own version, first seen Nov 8, 2026
- zxyycom/skillstheir own version, first seen Nov 8, 2026
- fang2hou/skillstheir own version, first seen Nov 1, 2026
- umararshad4/honey-stacktheir own version, first seen Nov 1, 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 versions18 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 18
- Agenta-AI/agenta★ 4.9ksame name
- adamentwistle/fable-skills★ 98same name
- nahid-sparktales/agent-dispatcher★ 59same name
- shabaraba/vibing.nvim★ 13same name
- tlzmw001/aitest-kit★ 13same name
- yasunori0418/skills★ 9same name
- umararshad4/honey-stack★ 5same name
- Lixiuxiu559/AgentForge★ 4same name
- Gusen1453/creed★ 3same name
- adolfoaranaes12/BMAD-ENHANCED★ 2same name
8 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 repository21 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- code-reviewSkill
- deliverSkill
- diagnoseSkill
- engineering-docsSkill
- engineering-writingSkill
- evaluateSkill
- implementSkill
- legal-writingSkill
- natural-writingSkill
- orchestrateSkill
- refactorSkill
- releaseSkill
- retrospectSkill
- scopeSkill
- sessionSkill
- setup-repoSkill
- skill-craftSkill
- trackerSkill
- ui-ux-designSkill
The file
We split this skill into the rules it gives an agent
3 rules. 0 are common, 3 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 12 edited versions we read keep these.
3
Distinctive
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 13
Use one contract test per public behaviour, property tests for invariants, golden tests for stable wire formats, and end-to-end tests for a …
- Distinctive: line 14
Search existing tests first and extend one when appropriate.
- Distinctive: line 20
Run focused tests locally and the project's exact-head CI gates.
“Common” is measured: kept by at least 6 in 10 of the 12 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
Test design
Choose evidence for observable behaviour and keep the suite economical. For a bug or core logic, write the spec-derived failing test first and witness the reported failure. For substantial work, have an independent author derive acceptance tests from the approved spec. Deleting and trimming tests is normal work. When a test fails on your change, decide whether it caught a defect or only pins the implementation: fix the code in the first case, and rewrite or delete the test in the second. Do not loosen an assertion that encodes a spec clause your change breaks; raise the conflict with the scope owner. Report every removed or weakened test and its reason.
Choose and prune
- Make illegal states unrepresentable where possible. Do not test states excluded by types, schemas or lint; identify the enforcing check.
- Use one contract test per public behaviour, property tests for invariants, golden tests for stable wire formats, and end-to-end tests for a few real paths. Choose the lowest boundary that proves the behaviour.
- Search existing tests first and extend one when appropriate. Budget at most one new test per new behaviour; name the plausible defect only it catches.
- On touch, delete duplicates, type-impossible cases and prose assertions; merge overlapping cases and rewrite tests coupled to implementation at the public boundary. A test that fails on a pure refactor needs repair.
- Mock only process, network, clock and randomness edges. Use real owned collaborators. See boundaries and judgement contrasts.
Apply the safeguards in judgement whenever you add, keep or prune tests.
Run focused tests locally and the project's exact-head CI gates. For risky logic, run mutation testing on the changed surface and investigate survivors. Report net tests added and removed in the PR, and identify unverified evidence.
How it runs
For substantial work, use provenant run --use review.tests --file review.md --explain with a pinned writer commit. Arrange an independent acceptance-test author separately.
For pruning, use explore over crates, then batch with disjoint writer worktrees and write_scope bindings into one integration branch (orchestrate). Inspect provenant run --use batch --file batch.toml --explain before submission.
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.

