Nw Design MethodologySAFE
AI agents that guide you from idea to working code, with you in control at every step.
Overview
AI agents that guide you from idea to working code, with you in control at every step.
1a864ec81cdeOBSERVED · 2026-10-14What it tells the agent
The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.
---
name: nw-design-methodology
description: Apple LeanUX++ design workflow, journey schema, emotional arc patterns, and CLI UX patterns. Load when transitioning from discovery to visualization or when designing journey artifacts.
user-invocable: false
disable-model-invocation: true
---
# Design Methodology (Apple LeanUX++)
## Design Workflow
```
PHASE 1 PHASE 2 PHASE 3 PHASE 4
Journey Mapping Emotional Design TUI Prototyping Integration Check
| | | |
v v v v
"What's the flow?" "How should it feel?" "What does it look?" "Does it connect?"
```
### Phase 1: Journey Mapping (1-2 days)
- Techniques: User journey mapping | goal-completion flow | step identification
- Question: "What complete journey is the user trying to accomplish?"
- Output: Journey map with steps, commands, and touchpoints
### Phase 2: Emotional Design (1 day)
- Techniques: Emotional arc design | form follows feeling | transition analysis
- Question: "How should the user FEEL at each step?"
- Output: Emotional annotations on journey map
### Phase 3: TUI Prototyping (1-3 days)
- Techniques: Progressive fidelity | ASCII mockups | TUI design patterns
- Question: "What does each step look like?"
- Output: TUI mockups for each journey step
### Phase 4: Integration Check (1 day)
- Techniques: Shared artifact tracking | horizontal coherence | CLI vocabulary
- Question: "Do all pieces connect properly?"
- Output: Validated journey with integration checkpoints
## Journey Schema
```yaml
schema_version: 1
journey:
name: "{Goal Name}"
goal: "{What user is trying to accomplish}"
persona: "{User persona reference}"
emotional_arc:
start: "{Initial emotional state}"
middle: "{Journey emotional state}"
end: "{Final emotional state}"
steps:
- id: 1
name: "{Step Name}"
command: "{CLI command or action}"
tui_mockup: |
+-- Step N: {Name} -----------------------------------------+
| {ASCII representation of CLI output} |
| ${variable} <-- tracked artifact |
+------------------------------------------------------------+
shared_artifacts:
- name: "{artifact_name}"
source: "{single source of truth file}"
displayed_as: "${variable}"
consumers: ["{list of places this appears}"]
emotional_state:
entry: "{How user feels entering step}"
exit: "{How user feels after step}"
integration_checkpoint: |
{What must be validated before proceeding}
failure_modes:
- "{What can go wrong at this step — used by DISTILL for error scenario generation}"
- "{Another failure scenario}"
gherkin: |
Scenario: {Step description}
Given {precondition}
When {action}
Then {observable outcome}
And shared artifact "${variable}" matches source
integration_validation:
shared_artifact_consistency:
- artifact: "{name}"
must_match_across: [1, 2, 3]
failure_message: "{Integration error description}"
changelog:
- date: "{YYYY-MM-DD}"
feature: "{feature-id}"
change: "{What changed in this update}"
```
## Emotional Arc Patterns
### Confidence Building
Start: Anxious/Uncertain | Middle: Focused/Engaged | End: Confident/Satisfied
Use when: Complex multi-step operations
### Discovery Joy
Start: Curious | Middle: Exploring | End: Delighted
Use when: Learning new features
### Problem Relief
Start: Frustrated | Middle: Hopeful | End: Relieved
Use when: Fixing issues or debugging
### Transition Rules
- Build confidence progressively through small wins
- Provide clear feedback at each step
- Error states guide to resolution rather than adding frustration
- Positive-to-negative transitions need explicit warning or buffer step
## Apple Design Principles Applied
- **Form Follows Feeling**: Design for emotion first, function second
- **Concentrated Focus**: One thing done excellently beats many done adequately
- **Material Honesty**: Respect the medium -- CLI should feel like CLI
- **Hidden Quality**: Excellence in details users may never see
## CLI UX Patterns (clig.dev)
### Command Structure
- Pattern: `tool [noun] [verb]` or `tool [verb] [noun]`
- Pick one pattern consistently across entire journey
- Example: `crafter agent create` or `crafter create agent`
### Feedback Principles
Responsive: print something in <100ms | Progress: show for long operations | Transparent: show what is happening | Recoverable: clear errors with suggested fixes
### Progressive Disclosure
- Level 1 (default): Basic output for common use
- Level 2 (--verbose): Detailed output for power users
- Level 3 (--debug): Diagnostic output for troubleshooting
### Help Design
Implement --help on every command | Make help discoverable | Provide contextual suggestions
## Output Formats
Three artifact types produced:
1. **Visual Journey** (`journey-{name}-visual.md`): ASCII flow diagram with emotional annotations and TUI mockups per step
2. **Structured Schema** (`journey-{name}.yaml`): Machine-readable journey definition following schema above
3. **Gherkin Scenarios** (`journey-{name}.feature`): Testable acceptance scenarios from each journey step
All artifacts go to `docs/feature/{feature-id}/discuss/`.Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | NA |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (0)
No findings outside the package's declared scope.
Gates applied: no_behavioural_pass.
1a864ec81cdefull audit observations/trust-audit/skill/nwave-ai__nw-design-methodology.json · Report an issue / request a re-scanAudit history
Every audit this skill has had.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-14 | 1a864ec81cde | SAFE | B | 89 | first audit |
Questions
What does the Nw Design Methodology skill do?
AI agents that guide you from idea to working code, with you in control at every step.
Is Nw Design Methodology safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can Nw Design Methodology access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
How current is this page?
The grade is for one exact copy of the source (1a864ec81cde), read on 2026-10-14. The repository is watched, and a new audit runs when it changes — this is the first audit.