persona-test
Skill0 starsChanged 7 months agoMITView on GitHub
Test any design, copy, UX, or product decision against simulated user personas. Use when the user wants to compare options, test headlines, evaluate button labels, pick between designs, grade copy, or get feedback on any product decision from multiple perspectives. Triggers on phrases like "test this", "which is better", "compare these", "run against personas", "persona test", "grade these options", "which copy works", "test this headline", "which version", or any A/B/C comparison of creative or product options.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- iamneilroberts/claude-skillstheir own version, first seen Jul 3, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- iamneilroberts/claude-skills★ 26same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository7 other files it gives its agents.
- browseSkill
- debriefSkill
- map-featureSkill
- progressive-summarySkill
- prompt-engineerSkill
- screen-reviewSkill
- spec-kitSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 6 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈6
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Test any product decision against 10 diverse simulated personas.
- Distinctive: line 16
If the user provides specific candidates to compare, use those directly.
- Distinctive: line 18
If the user provides a topic without candidates (e.g., "headline for a productivity app"), research the domain first:
- Distinctive: line 19
Search for best practices, frameworks, and examples in that domain
- Distinctive: line 27
Test each candidate against the persona panel.
- Distinctive: line 46
When simulating each persona, commit to their perspective.
- Very specific: line 72
After presenting results, offer to save them.
≈ “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
Persona Test
Test any product decision against 10 diverse simulated personas. Works for copy, headlines, UI patterns, feature names, pricing, onboarding flows, button labels — anything where you want signal from multiple perspectives before committing.
How it works
The skill runs in three phases: Generate, Test, Rank.
Phase 1: Generate (or accept) candidates
If the user provides specific candidates to compare, use those directly.
If the user provides a topic without candidates (e.g., "headline for a productivity app"), research the domain first:
- Search for best practices, frameworks, and examples in that domain
- Generate 25 candidate options across varied approaches (benefit-first, identity-first, metaphor, contrast, provocative, etc.)
- Select the top 15 for testing based on variety and quality
The research step matters — candidates informed by real copywriting/design research outperform random brainstorming.
Phase 2: Test against personas
Test each candidate against the persona panel. For each persona × candidate combination, collect:
- A 1-10 rating on each dimension
- A 1-sentence gut reaction in the persona's voice
Default personas (10, chosen for demographic and professional diversity):
| # | Persona | Why they're useful |
| 1 | 25yo creative freelancer | Tech-comfortable, design-aware, ad-skeptical |
| 2 | 35yo working mom, 2 kids | Time-starved, practical, low patience for jargon |
| 3 | 42yo startup founder | Ambitious, pattern-matching against other products, premium-sensitive |
| 4 | 28yo teacher | Moderate tech comfort, values clarity, limited budget |
| 5 | 50yo corporate executive | High expectations, jargon-tolerant in business context, premium-oriented |
| 6 | 22yo recent grad | Digital native, trend-aware, budget-conscious |
| 7 | 38yo small business owner | Pragmatic, ROI-focused, overwhelmed by options |
| 8 | 30yo software engineer | Technical, skeptical of marketing, values substance |
| 9 | 45yo real estate agent | Relationship-focused, mobile-first, wants tools that save time |
| 10 | 33yo nurse | Shift worker, high stress, needs things to just work |
When simulating each persona, commit to their perspective. A nurse doesn't think about "systems" the same way a software engineer does. A working mom has different patience thresholds than a recent grad. The gut reactions should sound like real people, not marketing evaluations.
Default dimensions (3):
| Dimension | What it measures |
| Clarity | Could this person understand what's being offered in under 3 seconds? |
| Desire | Does this make them want to continue / buy / engage? |
| Premium Feel | Does this feel high-quality, trustworthy, and worth their time? |
Phase 3: Rank and report
Compile results into:
- Ranked table — all candidates sorted by total average score, showing per-dimension averages
- Dimension winners — which candidate scored highest on each individual dimension
- Top 5 deep dive — for the top 5 candidates, include:
- Full score breakdown
- Notable persona reactions (especially disagreements — bimodal distributions are important signal)
- Strengths and weaknesses
- Which audience this candidate works best for
- Strategic recommendations — suggest which candidate to use in which context (e.g., "Use X for Product Hunt launch, Y for mass-market landing page, Z for in-app")
- Key insights — patterns that emerged across the testing (e.g., "personas over 40 consistently preferred concrete language over metaphors")
Saving results
After presenting results, offer to save them. Ask the user where they'd like the results saved, or save to the current project directory as persona-test-results/[Topic]-[Date].md.
Include the full ranked table, top 5 analysis, and strategic recommendations.
Arguments
The skill accepts these optional arguments:
- Topic (positional): What you're testing. E.g.,
"headline for welcome screen" - --candidates: Comma-separated list of specific options to test. Skips the generation phase.
- --dimensions: Comma-separated custom dimensions. E.g.,
"Trust, Excitement, Clarity" - --personas: Comma-separated custom personas. E.g.,
"tech founders, enterprise buyers, solo creators". Brief descriptions are expanded into full personas internally. - --count: Number of candidates to generate (default 25, top 15 tested)
- --top: Number of candidates for the deep-dive analysis (default 5)
Examples
/persona-test "headline for welcome screen"Researches headline best practices, generates 25, tests top 15, ranks, saves.
/persona-test --candidates "Connect, Enable, Unlock, Turn On" "button label for granting permissions"Tests the 4 provided options directly, no generation phase.
/persona-test "onboarding copy tone" --dimensions "Warmth, Trust, Clarity, Motivation" --personas "privacy-conscious millennial, busy executive, first-time app user, skeptical engineer"Custom dimensions and personas.
/persona-test "pricing page: monthly vs annual toggle placement"Researches pricing page UX, generates layout/copy candidates, tests.
What makes this skill different from just asking "which is better?"
Three things:
- Structured diversity. Ten personas with different jobs, ages, and contexts catch blind spots that a single perspective misses. A headline that resonates with a startup founder might alienate a nurse.
- Quantified signal. Gut feelings become comparable numbers. You can see that Option A wins on clarity but loses on premium feel — that's actionable in a way "I like A better" isn't.
- Bimodal detection. The most important finding is often disagreement. If a headline scores 9 with founders and 4 with teachers, that's not an average-of-6.5 situation — it's a targeting decision. The per-persona breakdown reveals these splits.
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.

