prototype-hci-pack
Skill19 starsChanged 6 days agoMITView on GitHub
Create a full HCI review pack for a prototype: conceptual model, state model, journeys, glossary, information architecture, consistency audit, failure path audit, heuristic evaluation, cognitive walkthrough, scorecard, and summary.
In this repository67 other files it gives its agents.
- accessibility-precheckSkill
- cognitive-friction-detectorSkill
- cognitive-walkthrough-beadsSkill
- cognitive-walkthroughSkill
- conceptual-model-beadsSkill
- conceptual-modelSkill
- consistency-audit-beadsSkill
- consistency-auditSkill
- error-and-time-metrics-analyzerSkill
- failure-path-audit-beadsSkill
- failure-path-auditSkill
- field-study-plannerSkill
- heuristic-eval-beadsSkill
- heuristic-evalSkill
- high-fidelity-prototype-plannerSkill
- information-architecture-beadsSkill
- information-architectureSkill
- interaction-type-selectorSkill
- journey-map-beadsSkill
- journey-mapSkill
- low-fidelity-prototype-plannerSkill
- mental-model-checkerSkill
- persona-synthesizerSkill
- prototype-hci-pack-beadsSkill
- prototype-hci-packSkill
- representative-task-writerSkill
- satisfaction-questionnaire-writerSkill
- state-model-beadsSkill
- state-modelSkill
- task-analysis-writerSkill
- usability-test-plannerSkill
- vocabulary-audit-beadsSkill
- vocabulary-auditSkill
- wireframe-criticSkill
- accessibility-precheckSkill
- cognitive-friction-detectorSkill
- cognitive-walkthrough-beadsSkill
- cognitive-walkthroughSkill
- conceptual-model-beadsSkill
- conceptual-modelSkill
- consistency-audit-beadsSkill
- consistency-auditSkill
- error-and-time-metrics-analyzerSkill
- failure-path-audit-beadsSkill
- failure-path-auditSkill
- field-study-plannerSkill
- heuristic-eval-beadsSkill
- heuristic-evalSkill
- high-fidelity-prototype-plannerSkill
- information-architecture-beadsSkill
- information-architectureSkill
- interaction-type-selectorSkill
- journey-map-beadsSkill
- journey-mapSkill
- low-fidelity-prototype-plannerSkill
- mental-model-checkerSkill
- persona-synthesizerSkill
- prototype-hci-pack-beadsSkill
- representative-task-writerSkill
- satisfaction-questionnaire-writerSkill
- The first 60 shown.
The file
Show its rules and text
We split this skill into the rules it gives an agent
20 rules. About 0 are common, 16 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈16
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 11
Use ultrathink. Work from the current repository, current files, current screenshots, current docs, and current prototype state.
- Distinctive: line 15
Ensure conceptual coherence across screens, flows, actions, and language.
- Distinctive: line 17
Leave durable artifacts that can guide implementation.
- Very specific: line 24
Write them to docs/hci/ unless the user asked for another path.
- Distinctive: line 38
Create the directory if it does not exist.
- Very specific: line 50
Use the conceptual-model skill logic:
- Distinctive: line 52
create a compact state model in markdown or Mermaid if useful
- Distinctive: line 60
create Mermaid state diagrams
- Distinctive: line 61
check UI state coverage: does every state have a distinct visual treatment?
- Distinctive: line 84
Create a glossary of canonical nouns, verbs, statuses, and labels.
- Very specific: line 94
Create a hierarchical sitemap of all screens/pages/views
- Distinctive: line 95
Document the navigation model: global, secondary, tertiary, escape hatches
- Distinctive: line 97
Check action placement: is each action near the objects it acts on?
- Distinctive: line 113
Create a table with:
- Distinctive: line 130
Create a summary table: category, items audited, handled, partial, not handled, critical gaps.
- Distinctive: line 159
After acting, will the feedback make sense?
- Very specific: line 164
Create hci-scorecard.md rating the prototype 1-5 on each dimension:
- Distinctive: line 193
Prefer recognition over recall.
- Distinctive: line 194
Prefer stable conceptual models over screen-by-screen patching.
- Distinctive: line 196
If the prototype is incomplete, still evaluate the structure that exists.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Build a full solo-review pack for this prototype or interface area:
$ARGUMENTS
Use ultrathink. Work from the current repository, current files, current screenshots, current docs, and current prototype state. Do not wait for user research unless absolutely blocked. If something is unknown, make the smallest reasonable assumption and label it clearly.
Goals
- Reduce confusion before user testing.
- Ensure conceptual coherence across screens, flows, actions, and language.
- Surface the highest-risk UX/HCI problems early.
- Leave durable artifacts that can guide implementation.
Visual default for new high-fidelity prototypes
Use Apple-inspired Liquid Glass on navigation and controls when a new high-fidelity prototype has no established design system. Use solid surfaces for readable content and dense data. Keep the border highlight optional, preserve text contrast and keyboard focus, and provide an opaque fallback for reduced transparency. Respect the product’s design system, touch input and reduced motion. Evaluate the treatment against the task and the Quantitative HCI example.
Produce these outputs
Write them to docs/hci/ unless the user asked for another path.
docs/hci/conceptual-model.mddocs/hci/state-model.mddocs/hci/user-journeys.mddocs/hci/glossary.mddocs/hci/information-architecture.mddocs/hci/consistency-audit.mddocs/hci/failure-path-audit.mddocs/hci/heuristic-evaluation.mddocs/hci/cognitive-walkthrough.mddocs/hci/hci-scorecard.mddocs/hci/hci-summary.md
Create the directory if it does not exist.
Required process
1) Understand the system
Extract and document:
- primary user types or actors
- primary objects/entities in the interface
- major actions users can take
- states for each important object
- system rules, permissions, and transitions
Use the conceptual-model skill logic:
- define users, objects, actions, states, rules
- create a compact state model in markdown or Mermaid if useful
- identify unclear or overloaded concepts
2) Define state models
For every stateful entity in the system:
- list all states with meaning and user-visible cue
- list allowed transitions with trigger, actor, and side effects
- list invalid transitions with explanation
- create Mermaid state diagrams
- check UI state coverage: does every state have a distinct visual treatment?
- flag ambiguous states, silent transitions, and undefined exits
3) Map the main journeys
Identify the 3 to 7 most important journeys for the scope. For each journey document:
- user goal
- entry point
- steps
- decision points
- error paths
- exit state
- likely confusion points
Include Mermaid flowcharts where useful. Prioritize:
- first-time user path
- core happy path
- edit/undo path
- failure/recovery path
- returning user path
4) Normalize language
Create a glossary of canonical nouns, verbs, statuses, and labels. For each term include:
- canonical term
- exact meaning
- where it appears
- terms to avoid or merge
Flag any concept drift such as one thing being called different names in different places.
5) Define information architecture
- Create a hierarchical sitemap of all screens/pages/views
- Document the navigation model: global, secondary, tertiary, escape hatches
- Analyze grouping logic and label clarity
- Check action placement: is each action near the objects it acts on?
- Measure depth and breadth: flag paths that are too deep or orphaned screens
6) Audit consistency
Review consistency across screens and flows:
- navigation placement
- page titles
- primary vs secondary actions
- destructive actions
- status presentation
- empty/loading/error/success states
- form layout and validation
- icon meaning
- modal, drawer, popover, and toast usage
- back/cancel/close behavior
Create a table with:
- item
- expected invariant
- observed variants
- risk
- recommendation
7) Audit failure paths
Review every non-happy-path state:
- empty states: does each empty screen explain why and offer a next step?
- loading states: is there always a loading indicator? timeout handling?
- validation errors: inline or summary? specific or vague? data preserved?
- permission failures: hidden, disabled, or error? explains how to fix?
- network failures: retry available? data preserved?
- partial completion: auto-saved? resumable?
- undo/cancel/recovery: confirmation dialogs? soft delete?
Create a summary table: category, items audited, handled, partial, not handled, critical gaps.
8) Run heuristic evaluation
Apply Nielsen-style heuristics rigorously:
- visibility of system status
- match between system and real world
- user control and freedom
- consistency and standards
- error prevention
- recognition rather than recall
- flexibility and efficiency of use
- aesthetic and minimalist design
- error recovery
- help/documentation
Output a severity-ranked issues table:
- ID
- heuristic violated
- where
- evidence
- why it matters
- severity (0 to 4)
- recommended fix
9) Run a cognitive walkthrough
For each critical journey, step through the interface and answer:
- Will the user form the right goal here?
- Will the user notice the correct action?
- Will the user understand that the action leads toward the goal?
- After acting, will the feedback make sense?
Log breakdowns with exact step references.
10) Score
Create hci-scorecard.md rating the prototype 1-5 on each dimension:
- Clarity of purpose
- Navigation clarity
- Terminology consistency
- State visibility
- Error prevention and recovery
- Recognition over recall
- Cross-screen consistency
- Failure path coverage
Include total score out of 40, top 3 strengths, top 3 weaknesses, and priority fixes.
For relevant interface patterns, use the Quantitative HCI catalogue at
docs/hci/models/interface-design-models.md in the Quantitative HCI repository.
Record the model ID, baseline, observed value, unit, sample size, test context,
and evidence. If a measurement is missing, record how to collect it. Keep
measurements alongside the scorecard's judgement scores.
11) Synthesize
Create hci-summary.md with:
- one-paragraph context recap
- top 5 confusion risks
- top 5 design invariants to enforce
- top 5 implementation priorities
- open assumptions and unknowns
Quality bar
- Be concrete, not generic.
- Refer to actual files, screens, components, routes, or flows where possible.
- Prefer recognition over recall.
- Prefer stable conceptual models over screen-by-screen patching.
- Distinguish structural issues from visual polish issues.
- If the prototype is incomplete, still evaluate the structure that exists.
Deliverable style
- Clear headings
- Markdown tables
- Mermaid diagrams where useful
- Direct language
- No filler
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.

