sharaf/product-experience-audit
Use when the user wants to audit a user journey, audit a signup/onboarding/checkout flow, do a UX audit, find the friction in a funnel, understand why users are dropping off or where they are being lost, or improve conversion in a web app — any diagnostic review of a multi-step, in-product flow. Use it whenever the user mentions drop-off, funnels, session replay, heatmaps, activation, time-to-value, cart or checkout abandonment, onboarding friction, or rage clicks, or wants to know where users struggle and what to fix first, even if they don't say "audit." Produces a severity-ranked, prioritized, experiment-validated improvement backlog via evidence-first intake, five parallel specialist lenses, and synthesis.
94
100%
Does it follow best practices?
Impact
72%
1.26xAverage score across 3 eval scenarios
Passed
No known issues
Evaluation results
82%
36%Onboarding Flow Audit: LoopSync B2B SaaS
Five-lens onboarding audit with journey brief and dark pattern detection
Journey brief present
25%
100%
Brief schema — 9 fields
12%
100%
Analytics marked missing
100%
100%
Five lenses addressed
37%
62%
Clean lens documented
0%
85%
Finding block template
50%
100%
Invite-step finding deduplicated
100%
100%
Dark pattern flagged for removal
100%
100%
Regulatory exposure cited
57%
100%
Falsifiable hypothesis
0%
12%
LIFT or Fogg model referenced
0%
0%
What's Working Well section
0%
100%
No analytics invented
100%
100%
53%
1%Checkout Funnel Audit
Checkout funnel audit with absolute-volume prioritization and analytics trust caveats
Absolute volume calculation
100%
100%
Correct top-priority step
100%
100%
Instrumentation sanity check
10%
0%
Analytics data-trust caveats
0%
0%
Mobile segment isolated
100%
100%
Hidden drop-offs flagged
0%
20%
Finding blocks with all 7 fields
0%
0%
Backlog alongside finding blocks
80%
20%
Needs-research labeling
60%
100%
What's Working Well section
40%
20%
Falsifiable hypothesis present
40%
80%
81%
7%Password-Reset Flow: Rapid Triage Report
Quick triage with ICE/PXL prioritization, no analytics
Triage scope stated
75%
62%
2-3 lenses selected
100%
12%
Lens justification present
100%
100%
Heuristic or qualitative lens included
37%
100%
9-field journey brief present
12%
37%
Missing evidence marked 'not provided'
87%
87%
ICE or PXL framework used
25%
100%
Per-factor scores shown
100%
100%
7-field finding blocks used
37%
100%
No fabricated drop-off rates
100%
87%
Magnitudes described as unmeasured
100%
100%
Treat benchmarks as directional
100%
66%
Next evidence step stated
100%
100%
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6