product-design-harness
Review Product Decisions Before Execution
Product teams can build quickly before testing their assumptions. This skill reviews user needs, evidence, and business viability to define a bounded next step.
Установить с помощью моего Агента
Скопируйте этот запрос в своего Агента. Он содержит каноническую страницу Skill и манифест.
Review the Skillstore skill "product-design-harness" from https://skillstore.io/skills/cis2042-product-design-harness.md and its manifest at https://skillstore.io/api/skills/cis2042-product-design-harness/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.Ваш Агент по-прежнему должен показать план и запросить все подтверждения, требуемые политикой безопасности.
Ресурсы для AI-агентов
Используйте эти ссылки, когда AI-агенту, crawler или script нужен чистый контекст вместо полной страницы.
Протестировать
Использование «product-design-harness». Review weekly report exports requested through support tickets, without evidence of managers' current reporting behavior.
Ожидаемый результат:
- Verdict: Verify. Weakest flow: Evidence.
- Missing proof: recurring reporting work and the cost of the current workaround.
- Next action: test a static report sample during workflow interviews.
- Boundary: no scheduled exports, live customer data access, or external delivery.
Использование «product-design-harness». Review a proposed product direction supported only by untraceable founder impressions.
Ожидаемый результат:
- Verdict: Stop/Reframe. Evidence tier: t0.
- Reason: no traceable signal supports the core problem.
- Reframed question: which specific actor experiences a recurring problem that can be observed?
- Boundary: research and reframing only; no implementation handoff.
Использование «product-design-harness». Prepare an implementation handoff from an approved continue review and its linked evidence receipts.
Ожидаемый результат:
- Handoff: context pack linked to the approved review.
- Scope: exact copy of the review's allowed actions.
- Controls: preserve approval gates, prohibited actions, unresolved assumptions, and stop conditions.
- Validation: check the review and context pack together before implementation.
Аудит безопасности
БезопасноAll 133 static findings are false positives involving documentation, schema identifiers, review terminology, or benign local validation logic. No evidence found of credential access, unauthorized data transfer, shell substitution, or audit-directed prompt injection in the reviewed context. Documented verdict overrides concern accountable product decisions, not permission to override security controls.
Факторы риска
📁 Доступ к файловой системе (10)
🌐 Доступ к сети (38)
Поделиться и цитировать этот отчет
Делитесь версионным отчетом об оценке, нейтральным значком, встраиваемой карточкой и цитатами. Skillstore публикует доказательства, не решая, безопасен ли этот Skill.
Копировать ссылку на отчёт
https://skillstore.io/skills/cis2042-product-design-harness/audits/3?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportЗначок Markdown
[](https://skillstore.io/skills/cis2042-product-design-harness?utm_source=security_passport_badge)Значок HTML
<a href="https://skillstore.io/skills/cis2042-product-design-harness?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/cis2042-product-design-harness/security.svg" alt="Skillstore security assessment" loading="lazy"></a>Встраиваемая карточка
<iframe src="https://skillstore.io/embed/skills/cis2042-product-design-harness.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>Академические ссылки (APA · BibTeX · CFF)
Цитата APA
cis2042. (2026). product-design-harness security audit report (audit version 3) [Author version unspecified]. Skillstore. https://skillstore.io/skills/cis2042-product-design-harness/audits/3Цитата BibTeX
@techreport{cis2042-cis2042-product-design-harness-2026,
author = {cis2042},
title = {product-design-harness security audit report (audit version 3)},
institution = {Skillstore},
year = {2026},
number = {3},
url = {https://skillstore.io/skills/cis2042-product-design-harness/audits/3},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "product-design-harness security audit report (audit version 3)"
version: "unspecified"
type: report
authors:
- name: "cis2042"
date-released: "2026-10-23"
url: "https://skillstore.io/skills/cis2042-product-design-harness/audits/3"
identifiers:
- type: other
value: "skillstore:cis2042-product-design-harness:audit:3"
description: "Skillstore immutable audit report identifier"
Оценка Skillstore
Почему такая оценка Достоверность доказательств: СреднийЧто вы можете построить
Assess a Feature Request
Review a proposed feature against user problems, supporting evidence, ownership cost, and the smallest useful test.
Challenge a Prototype Direction
Separate interview opinions from observed behavior and identify assumptions that require a focused prototype test.
Prepare a Bounded Agent Handoff
Validate an approved review and matching context pack before assigning implementation work to coding agents.
Попробуйте эти промпты
Review this product direction: [direction]. Current evidence: [sources]. Owner and organization: [details]. Select a mode and list missing intake. Do not implement.
Run Standard Gate for [feature] at [stage]. Use [brief and evidence]. Review all three flows independently. Return the canonical review with boundaries and stop conditions.
Run UX3 Council for [decision]. Use [brief, evidence receipts, and owner]. Decompose claims, review lanes independently, and complete challenge rounds. Red-team unanimous continue results before synthesis.
Use [continue review] to create a matching context pack. Copy scope and controls exactly. Validate both artifacts with the bundled checker. Report failures without implementing.
Лучшие практики
- Provide traceable evidence, counter-signals, affected stakeholders, and a named accountable owner before requesting a decision.
- Choose the smallest responsible review mode and keep human judgment separate from evidence tiers.
- Validate review artifacts and linked context packs together; review again when scope, data, risk, or stakeholders change.
Избегать
- Treating quotes, simulated users, or internal confidence as proof of real behavior.
- Averaging reviewer verdicts or removing counter-evidence to obtain approval.
- Turning a verify result into general implementation work or expanding an approved context pack.
Часто задаваемые вопросы
What does this skill review?
Which review mode should I use?
What outcomes can a review return?
Does the skill collect research automatically?
Do valid review artifacts prove that a decision is safe?
Can I use this with Claude, Codex, and Claude Code?
Сведения для разработчиков
Автор
cis2042Лицензия
MIT
Ревизия Skillstore
r3
Примечание о версии
Автор не указал версию.
Ссылка
3ee31ca64e16a02feb0f8db8f012ecc0de8a2c34
Актуальность поддержки
23.10.2026
Использование
5 загрузок · 0 просмотров
Структура файлов
📄 LICENSE
📁 agents/
📁 assets/
📁 docs/
📄 CODING-HARNESS-COMPARISON.md
📁 rfcs/
📄 UX3.md
📄 live-coding-context-pack.json
📄 live-coding-product-brief.json
📁 prompts/
📁 schemas/
📄 evidence-request.schema.json
📄 evidence-response.schema.json
📄 reviewer-verdict.schema.json
📄 uncertainty-review.schema.json
📄 verdict-override.schema.json
📁 scripts/
📄 SKILL.md