property-based-testing
Use this skill when you need to turn invariants, generation domains, and shrinking strategies into reviewable property-test candidates; triggers include 基于属性的测试 and property-based test design.
Install
npx skills add https://github.com/naodeng/awesome-qa-skills/tree/main/skills/en/testing-types/property-based-testing
claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install naodeng-awesome-qa-skills@llmmart
git clone https://github.com/naodeng/awesome-qa-skills.git
The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole naodeng/awesome-qa-skills collection as a plugin from our marketplace. Git is the plain clone.
Skill manifest
Property-Based Test Design
Turn invariants, generation domains, and shrinking strategies into reviewable property-test candidates. Produce PBT-## design candidates within the evidence boundary; do not execute tests or claim coverage or pass results.
When to Use
- Analyze domain invariants, input generation domains, constraints, failure examples, shrinking strategies, and existing properties.
- Preserve selection rationale, evidence gaps, priority, and validation actions.
- Inputs are incomplete but a bounded first pass can mark items unassessed or blocked.
Output Format Options
- Use Markdown by default; use tables, JSON, or CSV only when explicitly requested or required by the delivery format.
- Separate static analysis, unexecuted work, evidence states, and Human decisions; keep items unassessed, blocked, or NOT_RUN when runtime evidence is absent.
How to Use
- Read
prompts/property-based-testing.mdand provide the objective, scope, material, environment, and evidence. - Start with separate known, missing, conflicting, stale, out_of_scope, and assumptions entries.
- Produce PBT-## findings with source, evidence state, applicability, impact/priority, owner, close condition, and validation.
- Separate facts, evidence-backed inferences, recommendations, and Human decisions.
- Recommend follow-up validation without claiming execution.
Domain Execution Essentials
- Extract the input domain, sourced invariants, preconditions and observable oracles; do not infer properties solely from a function name or preference.
- Deliver domain/properties/preconditions, replay data and minimal-counterexample oracles; sorted output losing elements violates such a sorting contract.
- When deriving domain candidates, read and follow prompts/property-based-testing.md for the domain procedure and oracle; retain valid controls, evidence gaps, and validation actions for each item.
Core Constraints
- Do not invent invariants, generation domains, or shrink results, or treat a generator as proof of a finding.
- File presence, names, templates, and Eval configuration are not runtime evidence.
- Do not edit requirements, code, test assets, or target systems, or accept risk for a Human.
Pre-delivery Check
- The six-part input audit is complete.
- Every PBT-## has source, evidence state, applicability, concern, impact/priority, owner, close condition, and validation.
- Facts, inferences, recommendations, and Human decisions are separate.
- Unexecuted, unverified, unassessed, and pending-decision items are explicit.
Reference Files
- Read evals/eval.yaml and matching cases for regression; configuration does not prove project results.
- Use evals/trigger-prompts.csv and evals/local-rules.json for trigger checks; missing skill.selection evidence is BLOCKED.
Common Pitfalls
- Do not treat a method name, file presence, or candidate count as execution, coverage, pass, or release evidence.
- Do not fill missing invariants, generation domains, or results with convention; preserve unassessed, blocked, and pending items.
- Do not expand this specialist design into a complete strategy, full test cases, runtime execution, or a release decision.
Best Practices
- Complete the six-part input audit before selecting the smallest traceable and verifiable finding scope.
- Keep the source, evidence state, impact/priority, owner role, close condition, validation method, and residual risk for every finding.
- Write validation suggestions as next actions; do not upgrade package structure, candidate counts, or local Eval configuration into real quality conclusions.
Files (awesome-qa-skills)
-
agents
-
openai.yaml 339 B
Loading…
-
-
evals
-
cases
-
basic-success.yaml 831 B
Loading… -
domain-method-control.yaml 1.1 KB
Loading… -
domain-method-success.yaml 1.2 KB
Loading… -
edge-incomplete-input.yaml 815 B
Loading… -
edge-scope-boundary.yaml 915 B
Loading…
-
-
eval.yaml 526 B
Loading… -
local-rules.json 141 B
Loading… -
trigger-prompts.csv 673 B · in bundle
-
-
prompts
-
property-based-testing.md 3.5 KB
Loading…
-
-
SKILL.md 4.1 KB
--- name: property-based-testing description: Use this skill when you need to turn invariants, generation domains, and shrinking strategies into reviewable property-test candidates; triggers include 基于属性的测试 and property-based test design. license: PolyForm-Noncommercial-1.0.0 metadata: version: "1.0.0" compatibility: "" author: "Awesome QA Skills contributors" language: "en" category: "testing-types" source: "https://github.com/naodeng/awesome-qa-skills" license-url: "https://polyformproject.org/licenses/noncommercial/1.0.0" --- # Property-Based Test Design Turn invariants, generation domains, and shrinking strategies into reviewable property-test candidates. Produce PBT-## design candidates within the evidence boundary; do not execute tests or claim coverage or pass results. ## When to Use - Analyze domain invariants, input generation domains, constraints, failure examples, shrinking strategies, and existing properties. - Preserve selection rationale, evidence gaps, priority, and validation actions. - Inputs are incomplete but a bounded first pass can mark items unassessed or blocked. ## Output Format Options - Use Markdown by default; use tables, JSON, or CSV only when explicitly requested or required by the delivery format. - Separate static analysis, unexecuted work, evidence states, and Human decisions; keep items unassessed, blocked, or NOT_RUN when runtime evidence is absent. ## How to Use 1. Read `prompts/property-based-testing.md` and provide the objective, scope, material, environment, and evidence. 2. Start with separate known, missing, conflicting, stale, out_of_scope, and assumptions entries. 3. Produce PBT-## findings with source, evidence state, applicability, impact/priority, owner, close condition, and validation. 4. Separate facts, evidence-backed inferences, recommendations, and Human decisions. 5. Recommend follow-up validation without claiming execution. ## Domain Execution Essentials - Extract the input domain, sourced invariants, preconditions and observable oracles; do not infer properties solely from a function name or preference. - Deliver domain/properties/preconditions, replay data and minimal-counterexample oracles; sorted output losing elements violates such a sorting contract. - When deriving domain candidates, read and follow [prompts/property-based-testing.md](prompts/property-based-testing.md) for the domain procedure and oracle; retain valid controls, evidence gaps, and validation actions for each item. ## Core Constraints - Do not invent invariants, generation domains, or shrink results, or treat a generator as proof of a finding. - File presence, names, templates, and Eval configuration are not runtime evidence. - Do not edit requirements, code, test assets, or target systems, or accept risk for a Human. ## Pre-delivery Check - [ ] The six-part input audit is complete. - [ ] Every PBT-## has source, evidence state, applicability, concern, impact/priority, owner, close condition, and validation. - [ ] Facts, inferences, recommendations, and Human decisions are separate. - [ ] Unexecuted, unverified, unassessed, and pending-decision items are explicit. ## Reference Files - Read evals/eval.yaml and matching cases for regression; configuration does not prove project results. - Use evals/trigger-prompts.csv and evals/local-rules.json for trigger checks; missing skill.selection evidence is BLOCKED. ## Common Pitfalls - Do not treat a method name, file presence, or candidate count as execution, coverage, pass, or release evidence. - Do not fill missing invariants, generation domains, or results with convention; preserve unassessed, blocked, and pending items. - Do not expand this specialist design into a complete strategy, full test cases, runtime execution, or a release decision. ## Best Practices - Complete the six-part input audit before selecting the smallest traceable and verifiable finding scope. - Keep the source, evidence state, impact/priority, owner role, close condition, validation method, and residual risk for every finding. - Write validation suggestions as next actions; do not upgrade package structure, candidate counts, or local Eval configuration into real quality conclusions.
Comments (0)
Sign in to join the conversation.
Reviews (0)
No reviews yet.
No comments yet.