Persuasive UX Skill

Uses BJ Fogg's 7 Persuasive Technology Tools (from Captology) to audit a UI or flow and return a prioritized list of actionable UX improvements.

The 7 Tools — Quick Reference

| # | Tool | Core Mechanism | |---|------|----------------| | 1 | Reduction | Shrink effort — fewer steps, less friction, smart defaults | | 2 | Tunneling | Guided path — remove irrelevant choices, wizard-style progression | | 3 | Tailoring | Personalization — adapt content/UI to user context, goals, or history | | 4 | Suggestion | Kairos — surface the right prompt at the right moment | | 5 | Self-Monitoring | Real-time feedback — show users their progress or status | | 6 | Surveillance | Social visibility — peer awareness, leaderboards, "last active" | | 7 | Conditioning | Positive reinforcement — rewards, micro-celebrations, satisfying feedback |

Full definitions are in references/tools.md. Read it if you need depth on any tool.

---

Workflow

Step 1 — Understand the input

Accept any of the following input types:

• A text description of a feature, screen, or user flow
• A screenshot or mockup (analyze visually)
• An open-ended prompt like "improve this" or "why do users drop off here"

If the input is vague, ask ONE clarifying question: "What behavior are you trying to drive?" (e.g. sign up, complete a task, return more often). Don't ask more than one question.

Step 2 — Identify applicable tools

Mentally map the flow against all 7 tools. Ask:

• Where is effort or complexity creating friction? → Reduction
• Is the path unclear or branchy? → Tunneling
• Is the experience generic when it could be personal? → Tailoring
• Are suggestions poorly timed or missing entirely? → Suggestion
• Can users see their own progress? → Self-Monitoring
• Is there social context that could motivate? → Surveillance
• Are desired actions being rewarded? → Conditioning

Not every tool applies to every situation. Only surface what's genuinely relevant.

Step 3 — Output format

Return a short prioritized list (2–4 recommendations max). More is noise.

For each recommendation:

**[Tool Name]** — [One-line summary of the issue]
→ [Concrete change to make, specific to the input]
Why it works: [1–2 sentences of rationale]

Order by expected impact, highest first. If two tools have similar impact, prefer the one that requires less implementation effort.

---

Principles to keep in mind

• $B = MAP$ — Fogg's Behavior Model: Behavior happens when Motivation, Ability, and a Prompt converge. These 7 tools increase Ability (Reduction, Tunneling, Self-Monitoring) or sharpen the Prompt (Suggestion, Tailoring, Conditioning, Surveillance).
• Don't recommend all 7 — a focused list of 2–4 is more useful than exhaustive coverage.
• Be specific — "add a progress bar" beats "use self-monitoring". Tie every recommendation to the actual UI or flow described.
• Avoid dark patterns — Surveillance and Conditioning in particular can tip into manipulation. Flag if a recommendation risks feeling coercive.

---

Reference files

• references/tools.md — Full definitions, examples, and anti-patterns for all 7 tools. Read this when you need deeper context on a specific tool before recommending it.