ai-design-context 0.4.0

package/research/ux/high-risk-operational-flows.md

@@ -0,0 +1,89 @@
+---
+id: RESEARCH-00012
+alias: HIGH-RISK-OPERATIONAL-FLOWS
+slug: high-risk-operational-flows
+title: High-Risk Operational Flows
+object_type: research
+status: active
+version: 0.1.0
+category: ux
+tags:
+  - finance
+  - ai
+  - confirmation
+  - recovery
+  - trust
+maturity: validated
+risk_level: high
+last_reviewed_at: 2026-08-12
+relationships:
+  - type: derived_from
+    target: observations/finance/high-risk-wallet-recovery-flows.md
+  - type: derived_from
+    target: observations/finance/market-interface-adaptive-trading-surfaces.md
+  - type: derived_from
+    target: observations/ai/agent-consequential-action-controls.md
+  - type: related_to
+    target: RESEARCH-00011
+  - type: related_to
+    target: RESEARCH-00006
+---
+
+# High-Risk Operational Flows
+
+## Why Study It
+
+AI agents, wallets, trading surfaces, and account-security flows increasingly let users perform actions with real consequences. These surfaces need more than visual polish: they need risk-proportional confirmation, preserved context, interruption, recovery, and auditability.
+
+## Source Signals
+
+- Wallet product notes treat withdrawals, KYC, 2FA, passkeys, activation, and backend challenges as distinct recoverable flows.
+- Market-interface notes separate chart domain data from chart-engine data and adapt dense trading controls into mobile sheets or tablet modals instead of compressing desktop UI.
+- OpenAI's ChatGPT agent exposes user control through permission before consequential actions, interruption, takeover, stop, and task progress visibility.
+- OpenAI's Codex safety guidance uses sandbox boundaries, approval policies, network policy, and telemetry so higher-risk actions stop for review and remain inspectable.
+- Microsoft agent guidance frames agent UX as a full lifecycle from expectation setting to correction and recovery.
+- Microsoft HAX and Google PAIR both emphasize expectation setting, user control, correction, recovery, and failure planning for human-AI systems.
+
+Sources:
+
+- Wallet design-decision note captured on 2026-08-12.
+- Wallet product-requirement note captured on 2026-08-12.
+- Market-interface charting note captured on 2026-08-12.
+- https://openai.com/index/introducing-chatgpt-agent/
+- https://openai.com/index/running-codex-safely/
+- https://learn.microsoft.com/en-us/agents/design-guidelines/human-centered-design
+- https://www.microsoft.com/en-us/haxtoolkit/ai-guidelines/
+- https://pair.withgoogle.com/guidebook-v2/
+
+## Observed Design Behavior
+
+- Consequential actions are gated by the action's risk, reversibility, and side effects.
+- The object being acted on stays visible through confirmation and challenge states.
+- Recovery asks only for the missing or challenged input instead of restarting the whole task.
+- Distinct risk gates get distinct language and controls; generic error UI is avoided.
+- Users can interrupt, stop, retry, or take over long-running delegated work.
+- Mobile presentations preserve decision context with explicit sheet/modal contracts.
+- Audit and evidence trails matter when action can affect money, account access, external communication, or other people.
+
+## Design Takeaways For Agents
+
+- Classify actions by consequence before choosing the UI surface.
+- Preserve object, amount, recipient, source, scope, and side-effect context through every challenge state.
+- Make confirmations proportional: low-risk reversible actions may use lightweight undo; irreversible or external-side-effect actions need review before execution.
+- Keep the user's entered data stable while adding only the required challenge field.
+- Expose what will happen, what is still editable, what can be cancelled, and what evidence will remain.
+- For agentic work, combine lifecycle visibility with explicit approval for consequential actions.
+
+## Candidate Rule Directions
+
+- Confirm consequential actions with preserved context.
+- Make challenge and recovery states additive rather than restarting the user journey.
+- Treat mobile sheets for dense high-risk controls as contracts, not just responsive styling.
+
+## Do Not Copy Blindly
+
+- Do not add heavyweight confirmation to every harmless action.
+- Do not let a confirmation hide the object, amount, recipient, or side effects.
+- Do not expose internal backend codes as user-facing risk language.
+- Do not imply that AI approval logs or confirmations make unsafe actions acceptable.
+- Do not use this research as legal, compliance, trading, banking, or security policy.

package/research/ux/motion-as-state-continuity.md

@@ -0,0 +1,71 @@
+---
+id: RESEARCH-00010
+alias: MOTION-AS-STATE-CONTINUITY
+slug: motion-as-state-continuity
+title: Motion As State Continuity
+object_type: research
+status: draft
+version: 0.1.0
+category: ux
+tags:
+  - motion
+  - state
+  - continuity
+  - design-system
+maturity: seed
+risk_level: high
+relationships:
+  - type: derived_from
+    target: observations/interaction/figma-motion-as-system-capability.md
+  - type: derived_from
+    target: observations/interaction/chrome-scoped-view-transitions.md
+  - type: related_to
+    target: RESEARCH-00008
+  - type: related_to
+    target: RESEARCH-00005
+---
+
+# Motion As State Continuity
+
+## Why Study It
+
+Motion is often requested as generic polish after screens are assembled. Current tooling and platform capabilities instead make it possible to define motion as a reusable, inspectable part of state behavior.
+
+## Source Signals
+
+- Figma Motion places timing, easing, and keyframes beside components and variables and exposes them for implementation.
+- Element-scoped View Transitions can preserve page interactivity while explaining local state change.
+- Existing accessibility research requires equivalent function when non-essential motion is reduced.
+- Arc product research already identifies orientation and context preservation as legitimate motion purposes.
+
+Sources:
+
+- https://www.figma.com/blog/introducing-figma-motion/
+- https://developer.chrome.com/blog/element-scoped-view-transitions
+- https://developer.chrome.com/docs/css-ui/view-transitions/element-scoped-view-transitions
+
+## Observed Design Behavior
+
+- Motion connects a known start state to a known end state.
+- Local transitions preserve unrelated context and interactivity.
+- Reusable timing and easing values improve consistency and handoff.
+- Reduced-motion behavior preserves the same information, focus, and action outcome.
+
+## Design Takeaways For Agents
+
+- Specify the state relationship before duration or easing.
+- Reuse named motion roles rather than creating unrelated animation values.
+- Keep feedback transitions short enough that they do not delay expert work.
+- Define reduced-motion and unsupported-platform fallbacks in the same contract.
+- Verify focus, loading, error, and completion states without relying on movement alone.
+
+## Candidate Rule Directions
+
+- Use motion to explain state continuity, not to decorate.
+- Store repeated motion behavior as system roles with explicit fallbacks.
+
+## Do Not Copy Blindly
+
+- Do not treat every component as an animation opportunity.
+- Do not require a browser API without a progressive fallback.
+- Do not remove necessary feedback when reduced motion is active.

package/research/visual/expressive-system-ui-2026.md

@@ -0,0 +1,75 @@
+---
+id: RESEARCH-00009
+alias: EXPRESSIVE-SYSTEM-UI-2026
+slug: expressive-system-ui-2026
+title: Expressive System UI In 2026
+object_type: research
+status: draft
+version: 0.1.0
+category: visual
+tags:
+  - visual
+  - modern-ui
+  - visual-quality
+  - expressive-design
+  - hierarchy
+  - responsive
+maturity: seed
+risk_level: medium
+relationships:
+  - type: derived_from
+    target: observations/visual/apple-liquid-glass-functional-layer.md
+  - type: derived_from
+    target: observations/visual/material-3-expressive-system.md
+  - type: derived_from
+    target: observations/visual/chrome-content-first-adaptive-ui.md
+---
+
+# Expressive System UI In 2026
+
+## Why Study It
+
+AI-generated interfaces frequently converge on generic neutral cards or decorative effects without product hierarchy. Current platform guidance provides evidence for a more expressive direction, but also defines strong limits on where expression belongs.
+
+## Source Signals
+
+- Apple treats Liquid Glass as a functional controls and navigation layer, not a universal content material.
+- Material 3 Expressive combines typography, shape, color, component behavior, and motion through system roles.
+- Chrome's 2026 web UI guidance prioritizes content, reduces intrusive chrome, and adapts behavior to form factor.
+
+Sources:
+
+- https://developer.apple.com/design/human-interface-guidelines/materials
+- https://developer.apple.com/wwdc26/guides/design/
+- https://developer.android.com/develop/ui/compose/designsystems/material3
+- https://design.google/library/design-notes-material-3-expressive-liam-spradlin
+- https://developer.chrome.com/blog/new-in-web-ui-io26
+
+## Observed Design Behavior
+
+- The strongest materials and shapes identify functional hierarchy or meaningful emphasis.
+- Content remains readable and visually dominant beneath expressive controls.
+- Expression is encoded through reusable system roles rather than per-screen decoration.
+- Responsive behavior adapts density, placement, and input behavior instead of only scaling dimensions.
+- Accessibility preferences can reduce transparency, contrast ambiguity, or motion without removing function.
+
+## Design Takeaways For Agents
+
+- Choose an expression level from product purpose and platform context.
+- Name the content layer, functional layer, focal accent, and quiet supporting surfaces.
+- Reuse semantic typography, shape, material, color, and motion roles.
+- Treat platform effects as conditional capabilities, not cross-platform defaults.
+- Test the hierarchy without decorative effects; content and primary action must remain understandable.
+
+## Candidate Rule Directions
+
+- Keep expressive materials subordinate to content and functional hierarchy.
+- Encode expression through semantic system roles.
+- Adapt density and interaction to form factor and input method.
+
+## Do Not Copy Blindly
+
+- Do not translate Liquid Glass into unrestricted glassmorphism.
+- Do not interpret expressive design as maximal visual intensity.
+- Do not hide required actions merely to reduce chrome.
+- Do not assume experimental browser capabilities are production-ready across targets.

package/reviews/GENERATED_INDEX.md

@@ -0,0 +1,9 @@
+# Generated Reviews Index
+
+This file is generated from registry metadata. Do not edit manually.
+
+| ID | Alias | Title | Category | Status | Maturity | File Path | Required Checklists | Validates |
+| --- | --- | --- | --- | --- | --- | --- | --- | --- |
+| REVIEW-00001 | REVIEW-TODO-BENCHMARK | Todo App Benchmark Reference Review | reference-eval | draft | seed | reviews/todo-app-benchmark-reference-review.md | CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist | REF-TODO-BENCHMARK / REF-00001 Todo App Benchmark Reference Project |
+| REVIEW-00002 | REVIEW-TODO-RUNNABLE-FIXTURE | Todo Runnable Fixture Review | reference-eval | draft | seed | reviews/todo-reference-fixture-review.md | CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist | REF-TODO-BENCHMARK / REF-00001 Todo App Benchmark Reference Project |
+| REVIEW-00003 | REVIEW-CONSEQUENTIAL-ACTION-CONFIRMATION | Consequential Action Confirmation Validation Review | reference-eval | active | validated | reviews/consequential-action-confirmation-review.md | CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist | REF-CONSEQUENTIAL-ACTION-CONFIRMATION / REF-00002 Consequential Action Confirmation Reference Project<br>HIGH-RISK-OPERATIONAL-FLOWS / RESEARCH-00012 High-Risk Operational Flows<br>UX-006 / RULE-00017 Confirm Consequential Actions With Preserved Context<br>PAT-007 / PAT-00007 Consequential Action Confirmation<br>PROMPT-CONSEQUENTIAL-ACTION-REVIEW / PROMPT-00006 Consequential Action Review Prompt |

package/reviews/README.md

@@ -0,0 +1,25 @@
+# Reviews
+
+Reviews are first-class feedback objects.
+
+Use reviews to evaluate research, rules, patterns, prompts, reference projects, and generated product output.
+
+## Object Format
+
+Review objects should follow `schema/review.schema.json` and be registered in `registry/objects.json` with typed relationships in `registry/relationships.json`.
+
+Use `templates/REVIEW_TEMPLATE.md` for new reviews.
+
+Use `reviews/GENERATED_INDEX.md` to discover migrated reviews. The Todo benchmark review is deliberately `NEEDS WORK`: it records evidence limits rather than implying that directional benchmark output is proof of product quality.
+
+## Graph Role
+
+```text
+observations -> research -> knowledge -> rules -> patterns -> prompts -> reference projects -> reviews
+```
+
+Reviews can validate objects, identify missing layers, or create new observations.
+
+## Contribution Rule
+
+Do not treat reviews as comments only. A review should produce traceable findings that can improve the knowledge graph.

package/reviews/consequential-action-confirmation-review.md

@@ -0,0 +1,92 @@
+---
+id: REVIEW-00003
+alias: REVIEW-CONSEQUENTIAL-ACTION-CONFIRMATION
+slug: consequential-action-confirmation-review
+title: Consequential Action Confirmation Validation Review
+object_type: review
+status: active
+version: 0.1.0
+category: reference-eval
+tags:
+  - review
+  - confirmation
+  - fintech
+  - agentic-ui
+last_reviewed_at: 2026-08-12
+maturity: validated
+risk_level: high
+relationships:
+  - type: requires
+    target: CHECK-00001
+  - type: implements
+    target: PROMPT-00006
+  - type: validates
+    target: REF-00002
+  - type: validates
+    target: RESEARCH-00012
+  - type: validates
+    target: RULE-00017
+  - type: validates
+    target: PAT-00007
+  - type: validates
+    target: PROMPT-00006
+---
+
+# Consequential Action Confirmation Validation Review
+
+## Verdict
+
+PASS for specification-level validation of `RESEARCH-00012`, `RULE-00017`, `PAT-00007`, and `PROMPT-00006`.
+
+This review does not validate a production wallet, trading system, legal policy, compliance process, backend risk control, or security boundary.
+
+## Scope
+
+Review the high-risk operational flow chain:
+
+- `research/ux/high-risk-operational-flows.md`
+- `rules/ux/UX-006.md`
+- `patterns/consequential-action-confirmation.md`
+- `prompts/CONSEQUENTIAL_ACTION_REVIEW.md`
+- `examples/consequential-action-confirmation-reference.md`
+
+## Evidence Reviewed
+
+- Wallet and market-interface product evidence captured on 2026-08-12.
+- Agentic-control research cited in `RESEARCH-00012`: OpenAI ChatGPT agent, OpenAI Codex safety, Microsoft agent design guidance, Microsoft HAX, and Google PAIR.
+- Local reference specification `REF-00002`.
+- Checklist coverage from `CHECK-00001`.
+
+## Confirmed Coverage
+
+- `UX-006` is traceable to upstream research and concrete local product observations.
+- `PAT-007` composes existing context preservation, textual recovery, and agent intervention rules instead of duplicating them.
+- `PROMPT-00006` gives agents a bounded review contract and explicitly forbids inventing legal, compliance, security, financial, trading, or backend policy.
+- `REF-00002` gives a concrete wallet withdrawal and agent-approval flow that exercises confirmation, challenge, loading, error, retry, success, cancellation, and audit boundaries.
+
+## Findings
+
+### Low — Validation is specification-level, not rendered
+
+The objects are valid for repository use as active design guidance, but the current evidence does not include a runnable fixture, screenshots, keyboard traversal, reduced-motion run, or production telemetry.
+
+### Low — Domain policy remains intentionally out of scope
+
+The guidance can shape UI confirmation and recovery contracts, but final policy for money movement, account security, fraud, compliance, trading, custody, and backend authorization must come from the product domain.
+
+## Missing Layers
+
+- Rendered wallet confirmation fixture.
+- Mobile bottom-sheet evidence for dense challenge controls.
+- Keyboard, focus recovery, reduced-motion, and error-copy screenshots.
+- Independent product/security review for any production fintech implementation.
+
+## Recommendations
+
+- Keep `RESEARCH-00012`, `RULE-00017`, `PAT-00007`, and `PROMPT-00006` at `active` / `validated`.
+- Do not promote these objects to `canonical` until a rendered reference fixture and independent review exist.
+- Use `PROMPT-00006` for wallet, security, destructive, publishing, and agent-side-effect flow reviews.
+
+## Handoff
+
+Next validation step: build a small rendered fixture for `REF-00002` and review mobile, desktop, keyboard, reduced-motion, confirmation, challenge, and retry evidence.

package/reviews/todo-app-benchmark-reference-review.md

@@ -0,0 +1,86 @@
+---
+id: REVIEW-00001
+alias: REVIEW-TODO-BENCHMARK
+slug: todo-app-benchmark-reference-review
+title: Todo App Benchmark Reference Review
+object_type: review
+status: draft
+version: 0.1.0
+category: reference-eval
+tags:
+  - review
+  - benchmark
+  - todo
+maturity: seed
+risk_level: high
+relationships:
+  - type: requires
+    target: CHECK-00001
+  - type: implements
+    target: PROMPT-00005
+  - type: validates
+    target: REF-00001
+  - type: cites
+    target: PROMPT-00003
+  - type: cites
+    target: evidence/todo/2026-06-25-codex-gpt-5/EVALUATION.md
+  - type: cites
+    target: evidence/todo/2026-06-25-codex-gpt-5/ai-design-rules/generated-output.md
+---
+
+# Todo App Benchmark Reference Review
+
+## Verdict
+
+NEEDS WORK
+
+## Scope
+
+Review `REF-00001` and its 2026-06-25 Todo benchmark evidence as a reference-project validation artifact.
+
+## Evidence Reviewed
+
+- `benchmarks/todo-app.md`
+- `evidence/todo/2026-06-25-codex-gpt-5/EVALUATION.md`
+- `evidence/todo/2026-06-25-codex-gpt-5/ai-design-rules/generated-output.md`
+- `PROMPT-00003` / Todo App Benchmark Prompt
+
+## Confirmed Coverage
+
+- The artifact explicitly names the `Task` object, Today surface, quick capture, optional detail, and failed-save recovery.
+- It cites all six registered patterns and the initial nine rules used by the benchmark run.
+- The output specifies empty, loading, error, mobile, keyboard-label, and touch-target intentions.
+
+## Findings
+
+### High — No rendered interaction evidence
+
+The evidence is a Markdown design artifact. It cannot verify actual keyboard focus, touch target geometry, reduced-motion behavior, layout stability, or recovery behavior in a running interface.
+
+Related: `CHECK-00001`, `A11Y-001`, `A11Y-002`, `A11Y-003`, `A11Y-004`, `PERF-001`.
+
+### High — Benchmark evaluation is not independent
+
+The same model generated and evaluated both outputs, and the baseline ran in a repository-aware conversation. The reported score difference is directional, not a reliable quality claim.
+
+Related: `REF-00001`, `PROMPT-00003`.
+
+### Medium — Performance claims are specified, not measured
+
+The artifact describes a stable loading shell but contains no runtime measurements, slow-network run, or layout-shift evidence.
+
+Related: `PERF-001`.
+
+## Coverage Gaps
+
+- No mobile or desktop screenshots.
+- No clean-session baseline.
+- No independent evaluator.
+- No keyboard, assistive-technology, or reduced-motion test execution.
+- No measured loading or layout stability evidence.
+
+## Required Follow-Up
+
+Run the Todo benchmark in a clean session, produce rendered mobile and desktop outputs, capture screenshots and interaction evidence, test keyboard and reduced-motion paths, measure loading behavior, and obtain an independent evaluation before promoting `REF-00001` to `validated`.
+
+The later runnable-fixture review records limited rendered evidence, but it does not validate the historical benchmark claim.

package/reviews/todo-reference-fixture-review.md

@@ -0,0 +1,77 @@
+---
+id: REVIEW-00002
+alias: REVIEW-TODO-RUNNABLE-FIXTURE
+slug: todo-reference-fixture-review
+title: Todo Runnable Fixture Review
+object_type: review
+status: draft
+version: 0.1.0
+category: reference-eval
+tags:
+  - review
+  - fixture
+  - visual-qa
+  - todo
+maturity: seed
+risk_level: medium
+relationships:
+  - type: requires
+    target: CHECK-00001
+  - type: implements
+    target: PROMPT-00005
+  - type: validates
+    target: REF-00001
+  - type: cites
+    target: examples/todo-reference/index.html
+  - type: cites
+    target: examples/todo-reference/review-evidence/2026-07-15/desktop-default.png
+  - type: cites
+    target: examples/todo-reference/review-evidence/2026-07-15/desktop-detail.png
+  - type: cites
+    target: examples/todo-reference/review-evidence/2026-07-15/mobile-error.png
+  - type: cites
+    target: examples/todo-reference/review-evidence/2026-07-15/mobile-detail.png
+---
+
+# Todo Runnable Fixture Review
+
+## Verdict
+
+NEEDS WORK for benchmark validation; observed fixture flows pass this limited rendered review.
+
+## Scope
+
+Review `examples/todo-reference/` at 1440×1000 and 390×844 against `CHECK-00001`.
+
+## Evidence Reviewed
+
+- Desktop default and desktop detail screenshots.
+- Mobile empty-submit error screenshot.
+- Mobile detail-sheet screenshot.
+- Playwright accessibility snapshots for names, `alert`, invalid input state, and focus transfer to the detail close control.
+
+## Confirmed Coverage
+
+- The first viewport prioritizes Today, task list, and quick capture over dashboard summaries (`PRD-001`, `IA-001`, `PAT-001`, `PAT-002`).
+- Empty submission preserves the input, marks it invalid, returns focus, and exposes textual recovery through an alert (`A11Y-002`, `A11Y-003`).
+- Mobile controls use 44px minimum dimensions and the primary capture action remains reachable (`UX-001`, `A11Y-001`, `PAT-003`).
+- Detail inspection keeps the desktop list visible and leaves part of the mobile list visible behind the sheet (`UX-003`, `PAT-005`).
+- The fixture uses semantic tokens and reserves the task surface during its simulated save state (`VIS-001`, `PERF-001`).
+
+## Findings
+
+### High — No paired rendered benchmark
+
+The reviewed fixture is not a clean baseline-vs-rules experiment. It cannot validate the `+2.09` benchmark difference or promote `REF-00001` beyond `draft` / `reviewed`.
+
+### Medium — Keyboard and reduced-motion paths are not executed end to end
+
+Visible focus and reduced-motion CSS exist, and the review captured focus transfer after an interaction. A full keyboard-only traversal and a browser run with `prefers-reduced-motion: reduce` were not captured.
+
+### Medium — No measured loading evidence
+
+The fixture simulates a save state, but no slow-network run, layout-shift metric, or runtime performance measurement was recorded.
+
+## Handoff
+
+Create a clean paired benchmark run with both rendered outputs, then capture independent evaluator notes, keyboard/reduced-motion evidence, and performance measurements before setting `REF-00001` to `validated`.

package/rules/GENERATED_INDEX.md

@@ -0,0 +1,23 @@
+# Generated Rules Index
+
+This file is generated from registry metadata. Do not edit manually.
+
+| ID | Alias | Title | Category | Status | Maturity | File Path | Used By Patterns | Related Objects |
+| --- | --- | --- | --- | --- | --- | --- | --- | --- |
+| RULE-00001 | PRD-001 | Daily Actions First | product | draft | seed | rules/product/PRD-001.md | PAT-001 / PAT-00001 Daily Home Surface<br>PAT-003 / PAT-00003 Mobile Primary Action | derived_from: THINGS-3 / RESEARCH-00002 Things 3<br>derived_from: TELEGRAM / RESEARCH-00001 Telegram<br>derived_from: APPLE-REMINDERS / RESEARCH-00003 Apple Reminders<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist |
+| RULE-00002 | PRD-002 | Low-Friction Capture Before Organization | product | draft | seed | rules/product/PRD-002.md | PAT-002 / PAT-00002 Quick Capture<br>PAT-004 / PAT-00004 Progressive Detail | derived_from: APPLE-REMINDERS / RESEARCH-00003 Apple Reminders<br>derived_from: THINGS-3 / RESEARCH-00002 Things 3<br>derived_from: TELEGRAM / RESEARCH-00001 Telegram<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist |
+| RULE-00003 | IA-001 | One Stable Home Surface | ia | draft | seed | rules/ia/IA-001.md | PAT-001 / PAT-00001 Daily Home Surface<br>PAT-005 / PAT-00005 Context-Preserving Preview | derived_from: TELEGRAM / RESEARCH-00001 Telegram<br>derived_from: THINGS-3 / RESEARCH-00002 Things 3<br>derived_from: APPLE-REMINDERS / RESEARCH-00003 Apple Reminders<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist |
+| RULE-00004 | IA-002 | Define The Object Model Before Screens | ia | draft | seed | rules/ia/IA-002.md | PAT-005 / PAT-00005 Context-Preserving Preview<br>PAT-006 / PAT-00006 Object Status List | derived_from: LINEAR / RESEARCH-00004 Linear<br>derived_from: TELEGRAM / RESEARCH-00001 Telegram<br>derived_from: ARC / RESEARCH-00005 Arc<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist<br>related_to: UX-005 / RULE-00016 Expose Agent Work And Intervention |
+| RULE-00005 | UX-001 | Thumb-Zone Primary Actions | ux | draft | seed | rules/ux/UX-001.md | PAT-002 / PAT-00002 Quick Capture<br>PAT-003 / PAT-00003 Mobile Primary Action | derived_from: THINGS-3 / RESEARCH-00002 Things 3<br>derived_from: APPLE-REMINDERS / RESEARCH-00003 Apple Reminders<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist |
+| RULE-00006 | UX-002 | Progressively Disclose Power | ux | draft | seed | rules/ux/UX-002.md | PAT-001 / PAT-00001 Daily Home Surface<br>PAT-002 / PAT-00002 Quick Capture<br>PAT-004 / PAT-00004 Progressive Detail | derived_from: THINGS-3 / RESEARCH-00002 Things 3<br>derived_from: APPLE-REMINDERS / RESEARCH-00003 Apple Reminders<br>derived_from: TELEGRAM / RESEARCH-00001 Telegram<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist |
+| RULE-00007 | UX-003 | Preserve Context During Inspection | ux | draft | seed | rules/ux/UX-003.md | PAT-005 / PAT-00005 Context-Preserving Preview<br>PAT-006 / PAT-00006 Object Status List<br>PAT-007 / PAT-00007 Consequential Action Confirmation | derived_from: ARC / RESEARCH-00005 Arc<br>derived_from: LINEAR / RESEARCH-00004 Linear<br>derived_from: TELEGRAM / RESEARCH-00001 Telegram<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist<br>related_to: UX-004 / RULE-00015 Use Motion To Explain State Continuity<br>related_to: UX-005 / RULE-00016 Expose Agent Work And Intervention<br>related_to: UX-006 / RULE-00017 Confirm Consequential Actions With Preserved Context |
+| RULE-00008 | VIS-001 | Semantic Tokens Only | visual | draft | seed | rules/visual/VIS-001.md | PAT-004 / PAT-00004 Progressive Detail<br>PAT-006 / PAT-00006 Object Status List<br>PAT-001 / PAT-00001 Daily Home Surface | derived_from: LINEAR / RESEARCH-00004 Linear<br>derived_from: THINGS-3 / RESEARCH-00002 Things 3<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist<br>related_to: VIS-002 / RULE-00014 Keep Expression Subordinate To Content |
+| RULE-00009 | A11Y-001 | 44x44 Touch Targets | accessibility | draft | seed | rules/accessibility/A11Y-001.md | PAT-001 / PAT-00001 Daily Home Surface<br>PAT-002 / PAT-00002 Quick Capture<br>PAT-003 / PAT-00003 Mobile Primary Action<br>PAT-004 / PAT-00004 Progressive Detail<br>PAT-005 / PAT-00005 Context-Preserving Preview<br>PAT-006 / PAT-00006 Object Status List | derived_from: APPLE-REMINDERS / RESEARCH-00003 Apple Reminders<br>derived_from: THINGS-3 / RESEARCH-00002 Things 3<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist |
+| RULE-00010 | A11Y-002 | Textual Input Error Recovery | accessibility | draft | seed | rules/accessibility/A11Y-002.md | PAT-002 / PAT-00002 Quick Capture<br>PAT-007 / PAT-00007 Consequential Action Confirmation | derived_from: TEXTUAL-ERROR-RECOVERY / RESEARCH-00006 Textual Error Recovery<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt<br>related_to: UX-006 / RULE-00017 Confirm Consequential Actions With Preserved Context |
+| RULE-00011 | PERF-001 | Reserve Layout Space For Asynchronous Content | performance | draft | seed | rules/performance/PERF-001.md | PAT-002 / PAT-00002 Quick Capture | derived_from: RESERVED-LOADING-SPACE / RESEARCH-00007 Reserved Loading Space<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt |
+| RULE-00012 | A11Y-003 | Visible Keyboard Focus | accessibility | draft | seed | rules/accessibility/A11Y-003.md | PAT-002 / PAT-00002 Quick Capture | derived_from: KEYBOARD-FOCUS-AND-MOTION / RESEARCH-00008 Keyboard Focus And Interaction Motion<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt |
+| RULE-00013 | A11Y-004 | Reduce Non-Essential Interaction Motion | accessibility | draft | seed | rules/accessibility/A11Y-004.md | PAT-005 / PAT-00005 Context-Preserving Preview | derived_from: KEYBOARD-FOCUS-AND-MOTION / RESEARCH-00008 Keyboard Focus And Interaction Motion<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist<br>related_to: UX-004 / RULE-00015 Use Motion To Explain State Continuity<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt |
+| RULE-00014 | VIS-002 | Keep Expression Subordinate To Content | visual | draft | seed | rules/visual/VIS-002.md | PAT-001 / PAT-00001 Daily Home Surface | derived_from: EXPRESSIVE-SYSTEM-UI-2026 / RESEARCH-00009 Expressive System UI In 2026<br>related_to: VIS-001 / RULE-00008 Semantic Tokens Only<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt |
+| RULE-00015 | UX-004 | Use Motion To Explain State Continuity | ux | draft | seed | rules/ux/UX-004.md | PAT-005 / PAT-00005 Context-Preserving Preview | derived_from: MOTION-AS-STATE-CONTINUITY / RESEARCH-00010 Motion As State Continuity<br>related_to: UX-003 / RULE-00007 Preserve Context During Inspection<br>related_to: A11Y-004 / RULE-00013 Reduce Non-Essential Interaction Motion<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt |
+| RULE-00016 | UX-005 | Expose Agent Work And Intervention | ux | draft | seed | rules/ux/UX-005.md | PAT-007 / PAT-00007 Consequential Action Confirmation | derived_from: CONTEXTUAL-AGENTIC-INTERFACE / RESEARCH-00011 Contextual And Observable Agentic Interfaces<br>related_to: IA-002 / RULE-00004 Define The Object Model Before Screens<br>related_to: UX-003 / RULE-00007 Preserve Context During Inspection<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt<br>related_to: UX-006 / RULE-00017 Confirm Consequential Actions With Preserved Context<br>related_to: PROMPT-CONSEQUENTIAL-ACTION-REVIEW / PROMPT-00006 Consequential Action Review Prompt |
+| RULE-00017 | UX-006 | Confirm Consequential Actions With Preserved Context | ux | active | validated | rules/ux/UX-006.md | PAT-007 / PAT-00007 Consequential Action Confirmation | derived_from: HIGH-RISK-OPERATIONAL-FLOWS / RESEARCH-00012 High-Risk Operational Flows<br>related_to: UX-003 / RULE-00007 Preserve Context During Inspection<br>related_to: A11Y-002 / RULE-00010 Textual Input Error Recovery<br>related_to: UX-005 / RULE-00016 Expose Agent Work And Intervention<br>related_to: CHECK-DESIGN-QA / CHECK-00001 Design QA Checklist<br>related_to: PROMPT-PROTOTYPE-REVIEW / PROMPT-00002 Prototype Review Prompt<br>validates: REF-CONSEQUENTIAL-ACTION-CONFIRMATION / REF-00002 Consequential Action Confirmation Reference Project<br>validates: REVIEW-CONSEQUENTIAL-ACTION-CONFIRMATION / REVIEW-00003 Consequential Action Confirmation Validation Review |

package/rules/INDEX.md

@@ -0,0 +1,37 @@
+# Rules Index
+
+Rules are reviewable design instructions derived from research.
+
+Use this folder after `research/` and before `patterns/`.
+
+```text
+observations -> research -> knowledge -> rules -> patterns -> prompts -> reference projects -> reviews
+```
+
+## Rule Domains
+
+- `product/` - product intent, repeated actions, task priority.
+- `ia/` - information architecture, object models, navigation structure.
+- `ux/` - interaction flow, progressive disclosure, mobile behavior.
+- `visual/` - visual system decisions that support product clarity.
+- `accessibility/` - accessibility requirements for usable interaction.
+
+## Current Rules
+
+- `product/PRD-001.md` - daily actions first.
+- `product/PRD-002.md` - low-friction capture before organization.
+- `ia/IA-001.md` - one stable home surface.
+- `ia/IA-002.md` - define the object model before screens.
+- `ux/UX-001.md` - thumb-zone primary actions.
+- `ux/UX-002.md` - progressively disclose power.
+- `ux/UX-003.md` - preserve context during inspection.
+- `visual/VIS-001.md` - semantic tokens only.
+- `accessibility/A11Y-001.md` - 44x44 touch targets.
+
+## Contribution Rule
+
+Do not add a rule unless it can point back to research. If the source is only taste, keep it out of `rules/`.
+
+Future rule metadata should follow `docs/KNOWLEDGE_ENGINE.md`.
+
+Manual indexes are transitional. Future indexes should be generated from registry metadata.

package/rules/accessibility/A11Y-001.md

@@ -0,0 +1,58 @@
+---
+id: RULE-00009
+alias: A11Y-001
+slug: 44x44-touch-targets
+title: 44x44 Touch Targets
+object_type: rule
+status: draft
+version: 0.1.0
+category: accessibility
+tags:
+  - accessibility
+  - touch
+maturity: seed
+risk_level: high
+relationships:
+  - type: derived_from
+    target: RESEARCH-00003
+  - type: derived_from
+    target: RESEARCH-00002
+---
+
+# A11Y-001 44x44 Touch Targets
+
+Status: Draft
+Domain: Accessibility
+
+## Source Research
+
+- `research/products/apple-reminders.md`
+- `research/products/things-3.md`
+
+## Why
+
+Fast mobile actions fail when controls are too small to hit reliably. A product can look clean and still be difficult to use if repeated actions require precise tapping.
+
+## Rule
+
+Interactive touch targets must be at least 44x44 CSS pixels unless the platform has a stricter local standard.
+
+## Good
+
+- Primary and repeated controls meet the minimum target size.
+- Small icons sit inside a larger clickable area.
+- Adjacent controls have enough spacing to avoid accidental taps.
+- Disabled controls remain visually clear.
+
+## Bad
+
+- Icon-only buttons use the icon size as the hit area.
+- Dense lists place destructive and positive actions too close together.
+- Mobile controls shrink below readable and tappable size to fit more content.
+
+## Review
+
+- Inspect every interactive control on the mobile layout.
+- Measure the clickable area, not only the visible icon.
+- Check dense list rows, toolbars, and bottom navigation.
+- Confirm keyboard and screen reader labels still exist where relevant.