hashgraph-online/awesome-codex-plugins / plugins/alpoxdev/hypercore/skills/hyper-prd
hyper-prd
Verified[Hyper] Use this skill when the user wants a product requirements document for a web, iOS, Android, or desktop product - a new idea to size and write, an existing PRD to revise, or a PRD to review - and expects one `prd.md` file with goals, metrics, scoped stories, EARS requirements, acceptance scenarios, and a handoff block. It decides platform and size first, asks only the decisions the user alone can make within the size's question budget, marks unknowns inline, and stops at the document. Do not use it for user flows, wireframes, diagrams, HTML previews, technical design, task breakdowns, or implementation.
| File | Kind | Size |
|---|---|---|
| SKILL.ko.md | markdown | 22.9 KB |
| SKILL.md | skill md | 21.3 KB |
| assets/evals/hyper-prd-cases.jsonl | other | 12.9 KB |
| references/platforms/android.ko.md | markdown | 11.6 KB |
| references/platforms/android.md | markdown | 11.1 KB |
| references/platforms/desktop.ko.md | markdown | 8.9 KB |
| references/platforms/desktop.md | markdown | 8.6 KB |
| references/platforms/ios.ko.md | markdown | 9.6 KB |
| references/platforms/ios.md | markdown | 9.5 KB |
| references/platforms/web.ko.md | markdown | 10.3 KB |
| references/platforms/web.md | markdown | 10.1 KB |
| references/prd-template.ko.md | markdown | 25.0 KB |
| references/prd-template.md | markdown | 23.3 KB |
| rules/intake-and-questions.ko.md | markdown | 6.9 KB |
| rules/intake-and-questions.md | markdown | 6.7 KB |
| rules/requirements-writing.ko.md | markdown | 10.0 KB |
| rules/requirements-writing.md | markdown | 9.7 KB |
| rules/validation.ko.md | markdown | 8.7 KB |
| rules/validation.md | markdown | 8.2 KB |
rules/validation.md
View on GitHub# Validation
Use this rule before saving and in every `review`. It owns the per-mode completion evidence, the pre-save checklist, the blocking-question handling, and the optional no-context reader test.
## 1. Completion evidence per mode
| Mode | Evidence that the run is done | What is not evidence |
|---|---|---|
| `create` | `prd.md` saved at `.hyper/prd/<slug>/prd.md`, and every pre-save checklist item passes | A chat message describing the document |
| `update` | The changed sections, a change-history row with the date and what changed, and every checklist item passing | Rewriting the whole file while claiming a section edit |
| `review` | A pass/fail table per checklist item, with the failing item named and its location | Editing the file, or a prose verdict without a table |
`review` never edits. When a review finds a fix, it names the fix and stops; applying it is an `update`.
## 2. Pre-save checklist
Run each item over the saved text, not over the intention behind it.
| # | Check | Fails when |
|---|---|---|
| 1 | Every `FR-###` maps to a goal or a `P1`/`P2`/`P3` story | A requirement cites nothing, or cites only a non-goal |
| 2 | Every metric has a target, a measurement means, and a measurement time | One of the three is missing, or the target has no unit |
| 3 | Non-goals actually cut scope | A non-goal restates a goal, or could be removed without changing what is built |
| 4 | No vague word remains | `빠르게`, `쉽게`, `적절히`, `충분히`, `사용자 친화적`, `fast`, `easy`, `appropriate`, `robust`, `seamless` appear as a requirement |
| 5 | No implementation mechanism appears | A file path, function or type name, newly chosen library, or invented schema name is named outside an existing external constraint |
| 6 | Every story has a happy path and a failure-or-edge scenario citing `FR-###` ids | One path is missing, or a scenario cites no requirement |
| 7 | Every `A-###` has an evidence grade and a consequence | An assumption sits in the body as fact, or has no "if wrong" line |
| 8 | `[확인 필요: …]` appears at most 3 times outside the release gate tables | A fourth inline marker exists in the prose sections |
| 9 | No empty or `N/A` section remains | A heading has no content, or says `N/A`/`TBD` |
| 10 | The handoff block is complete | It lacks platforms, size, P1 scope, remaining blocking questions, or next-stage inputs |
| 11 | The meta block carries status, platforms, size, and the updated date | One field is missing, or the platform or size decision has no one-line reason |
| 12 | The must-not-change behavior is stated or explicitly declared absent | The change touches an existing flow and the section is missing |
| 13 | Every `[M]` and `[M-조건]` item of each decided platform appears in that platform's release gate table with a conclusion | A decided platform's checklist has an `[M]` or `[M-조건]` item with no row, or a row with no conclusion |
| 14 | No breakpoint is named after a device | A breakpoint reads `iPhone 375px`, `iPad`, `desktop 1440`, or any other device class instead of a content width in CSS pixels |
Item 13 counts items, not rows: an `[M-조건]` item that reads "not applicable" with the condition and its reason passes, and so does an `[M]` item marked `[확인 필요: <question>]` as long as the document stays `status: draft`. Item 8 counts only the inline markers in the prose sections: a `[확인 필요: <question>]` standing as a gate row's conclusion is excluded from the cap, and the document stays `status: draft` while any gate row is `[확인 필요]`. Item 14 applies to every width the document names, in the goals, the platform block, or a requirement. A breakpoint is written as the content width in CSS pixels plus what in the content needs it, for example `640px where the metric row wraps`; a device name in a breakpoint is a defect because the device list changes while the content width does not.
Fix a failure in place and re-run only the failing item. Do not re-run the whole checklist to feel finished.
## 3. Blocking questions
- Any `Q-###` with `Blocking: yes` and no answer keeps `status: draft`.
- A `draft` document is still saved and reported; it is not withheld and not reported as complete.
- The completion report names each remaining blocking question and who must answer it.
- A document may pass every checklist item and still be `draft`; the checklist measures the text, the status measures whether a scope decision is missing.
## 4. Optional no-context reader test
Run this only when a subagent is available and the user has not forbidden it. It is never a required step, and a skipped test is stated as skipped rather than passed.
1. Write five expected questions the document must answer for a build stage: what is being built, for whom, what is out of scope, how success is measured, and what decision is still open.
2. Give a fresh subagent only the saved `prd.md` and ask those five questions.
3. Count the answers the document supports. A question the document cannot answer from its own text is a defect in that section, not in the reader.
4. Fold the defects back into the document and re-run the affected checklist items.
Record the outcome as `reader test: passed <n>/5` or `reader test: skipped (<reason>)`. Never claim a pass that was not run.
## 5. Review output shape
```text
| # | Check | Result | Location / note |
|---|---|---|---|
| 1 | FR maps to goal or story | pass | |
| 2 | Metric target and measurement | fail | Section 5, FR-007 has no target |
| 13 | Platform `[M]` items in the release gate table | fail | Android block, A9 has no row |
| 14 | Breakpoints named by content | fail | Section 9, web block says "iPhone 375px" |
```
One row per checklist item, in checklist order. After the table, one line naming the blocking questions and one line stating that no file was edited.
## Sources
> Checked 2026-11-07.
| Claim | Source |
|---|---|
| Independent testability as a gate on every story | [spec-kit spec-template](https://raw.githubusercontent.com/github/spec-kit/main/templates/spec-template.md) - "Each user story/journey must be INDEPENDENTLY TESTABLE" |
| Removing a non-applicable section instead of writing N/A | [spec-kit specify](https://raw.githubusercontent.com/github/spec-kit/main/templates/commands/specify.md) - "When a section doesn't apply, remove it entirely (don't leave as 'N/A')." |
| A clarification cap as a readiness gate | [spec-kit specify](https://raw.githubusercontent.com/github/spec-kit/main/templates/commands/specify.md) - "Maximum 3 [NEEDS CLARIFICATION] markers total" |
| A no-context reader test before other readers | [anthropics doc-coauthoring](https://github.com/anthropics/skills) - "Test the doc with a fresh Claude (no context) to catch blind spots before others read it" |
| Assumptions kept out of body fact | [phuryn create-prd](https://github.com/phuryn/pm-skills) - "Don't invent. What you don't know goes to 7.4 Assumptions and open questions, clearly marked, not into the body as fact." |
| Counter-metrics as part of a measurable goal set | [TerminalSkills prd-writing](https://github.com/TerminalSkills/skills) - "Counter-metrics: [Metrics we don't want to hurt]" |
| The implementation-mechanism audit | [ai-conductor prd](https://github.com/jstoup111/ai-conductor) - "It MUST NOT name the new internal mechanism by which this feature is built." |
| Output proportional to the problem, so the checklist stays short | [Martin Fowler, SDD tool comparison](https://martinfowler.com/articles/exploring-gen-ai/sdd-3-tools.html) - "it quickly became clear that the workflow was like using a sledgehammer to crack a nut" |
| Breakpoints chosen by content instead of device class | [Responsive web design basics](https://web.dev/articles/responsive-web-design-basics) - "Don't define breakpoints based on device classes" |
| Every mandatory store item reaching the release gate | [Play Data safety](https://support.google.com/googleplay/android-developer/answer/10787469) - "Even developers with apps that do not collect any user data must complete this form" |
### Evidence grade
`LOCAL` - the fourteen checklist items and the review table shape are this package's design. The marker cap and the removal rule are quoted from the cited specification guidance; the reader test is quoted from the cited authoring guidance.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| brainstormingobra/superpowersYou MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation). | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |
| brand-guidelinesanthropics/skillsApplies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| frontend-designanthropics/skillsGuidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| mcp-builderanthropics/skillsGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK). | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |