tools/sage-claude-plugin/skills/ux-heuristic-review/SKILL.md
Evaluates implementation against usability heuristics — Nielsen's 10, Norman's principles, Krug's laws
npx skillsauth add xoai/sage ux-heuristic-reviewInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Evaluates the implementation against usability heuristics. Runs ALONGSIDE the
core quality-review skill at the review phase. When installed, this review
is mandatory — it cannot be skipped by the agent.
Evaluate the implementation against the four heuristics most likely to catch user-facing failures. Takes ~2 minutes per task.
Evaluate against all 10 Nielsen heuristics plus Norman's design principles. Takes ~10 minutes per major feature.
H2: Real-World Match — Does the interface use the user's language? Are icons and metaphors recognizable? Is information organized the way users think?
H4: Consistency — Are the same actions called the same thing everywhere? Do similar elements behave the same way? Does it follow platform conventions?
H6: Recognition over Recall — Are all available actions visible? Do forms show expected formats? Can users see previous selections when making new ones?
H7: Flexibility — Are there keyboard shortcuts for frequent actions? Can expert users bypass introductory steps? Does it remember preferences?
H8: Minimalism — Is there unnecessary text? Are visual elements serving a purpose? Is the visual hierarchy clear?
H10: Help — Is contextual help available for complex features? Are tooltips concise and useful?
In addition to Nielsen's heuristics, check:
Each finding gets a severity:
| Severity | Definition | Action | |----------|-----------|--------| | Catastrophic | User loses data or can't complete primary task | MUST fix before merge | | Major | User significantly confused or frustrated | SHOULD fix before merge | | Minor | User notices oddity but works around it | Fix in next iteration | | Cosmetic | Violates principle but minimal user impact | Note for polish |
## UX Heuristic Review: [feature name]
Date: [date]
Mode: [BUILD light / ARCHITECT full]
### Findings
1. [H9 - MAJOR] Error message on login form says "401 Unauthorized"
→ Should say: "Wrong email or password. Check your details and try again."
→ Fix: Update error handler in login-form.tsx
2. [H1 - CATASTROPHIC] No loading state on payment submission
→ User clicks "Pay" and sees no response for 3-5 seconds
→ Fix: Add loading spinner to submit button, disable during processing
### Summary
- Catastrophic: 1 (must fix)
- Major: 1 (should fix)
- GATE RESULT: FAIL (catastrophic finding blocks merge)
heuristic-evaluation.md — Full heuristic descriptions and checkpointsusability-principles.md — Krug's laws and Norman's principleserror-and-recovery-design.md — Error taxonomy for H5/H9 evaluationdevelopment
Branch-per-initiative git discipline for all delivery workflows. Defines branch naming by workflow, the propose-confirm creation protocol, dirty-tree and detached-HEAD handling, the always user-gated merge protocol, worktree support for parallel sessions, and abandonment cleanup. Activates only in git repositories — silently inactive everywhere else. Use when starting /build, /fix, /architect, or /build-x at Standard+ scope, when resuming an initiative, when offering a merge at a completion checkpoint, or when the user wants a second concurrent initiative.
development
Drives task-by-task execution from an approved plan with quality gates between each task. Reads the plan, finds the next incomplete task, dispatches implementation, validates, updates progress, and continues. Use after a plan is approved and the user says "go", "start building", "execute the plan", or "implement the feature".
testing
Preserves and restores context across agent sessions using plan file checkboxes as source of truth. Use when starting a new session, resuming previous work, ending a session, or when the user says "continue from last time", "what was I doing", or "save progress".
tools
Captures agent mistakes, corrections, and discovered gotchas so they are not repeated. Use when: (1) a command or operation fails unexpectedly, (2) the user corrects the agent, (3) the agent discovers non-obvious behavior through debugging, (4) an API or tool behaves differently than expected, (5) a better approach is found for a recurring task. Also searches past learnings before starting tasks to avoid known pitfalls. Activate alongside the sage-memory skill — they share the same MCP backend but serve different purposes (sage-memory = codebase knowledge, sage-self-learning = agent mistakes and gotchas).