Skip to content
Back to skills

Mk 028

ASecurity

Persona evidence review. Use when the task requires a checklist distinguishing supported items from invented assumptions. Apply this method to the specified inputs and verify the deliverable

  • 259 stars
  • 0 votes
  • 0 copies
  • 5 views
  • Added October 23, 2026
ai-agentsgo

Security analysis

A100/100

Pro scans all 3 files and shows the line behind each finding

Scanned October 23, 2026

npx -y skills add backtrue/thinkwithblack-skill --skill mk-028 --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Mk 028?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Mk 028
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/backtrue-mk-028/badge)](https://www.skillsdirectory.com/skills/backtrue-mk-028)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: mk-028
description: "Persona evidence review. Use when the task requires a checklist distinguishing supported items from invented assumptions. Apply this method to the specified inputs and verify the deliverable"
---

# MK-028 | Persona evidence review

For a regular chat, upload `packs/mk-028.md`. It includes the shared rules, this method, and required references

## Purpose

This skill delivers: **A checklist distinguishing supported items from invented assumptions**

Hypotheses are useful but do not establish demographics or market share; do not fabricate interview content or customer experiences

## Required inputs

Provide: Persona or need hypotheses to check, product facts, customer wording, survey or service records and sources

## Method

1. Treat Persona/need hypotheses as candidate answers initially; AI generation does not make them market facts
2. Use actual evidence such as interviews, search terms, customer-service records, and existing pages to check customer problems, searches, and next steps
3. If the AI invents an unheard-of problem, a page the company does not have, or an action customers may not want to take, mark insufficient evidence or the company's current inability to support it

## Calculations and interpretation

Treat customer roles and motivations as hypotheses initially; do not establish someone's actual identity from a single purchase behavior

## Required output fields

| Order | Field | How to complete |
|---:|---|---|
| 1 | Original Persona or need hypothesis | Use the supplied source, calculation, or explicit unknown |
| 2 | Source and location | Use the supplied source, calculation, or explicit unknown |
| 3 | Supported, contradicted or unknown | Use the supplied source, calculation, or explicit unknown |
| 4 | What the product can address | Use the supplied source, calculation, or explicit unknown |
| 5 | What it cannot address | Use the supplied source, calculation, or explicit unknown |
| 6 | Keep, revise or remove | Use the supplied source, calculation, or explicit unknown |
| 7 | Next evidence needed | Use the supplied source, calculation, or explicit unknown |

## Verification cases

**MK-028-C05 | An order contains maternity and baby products**

Expected: Do not immediately identify the buyer as a mother; separate the buying and using roles

**MK-028-C06 | Different clues support two situations**

Expected: Keep both candidates and request the data that would distinguish them

## Method sources

Source file: Google Ads 第一堂|看懂需求,才知道廣告怎麼接

Location: Hypothesis

Source basis: extracted from course materials

Source URL: https://docs.google.com/presentation/d/15ZOg0pX_xikpdA3FwGEoCSVgWortvCnxSe4hz4ugizg/edit

English translation of source excerpt:

> Use a hypothetical dehumidifier example to explain the event, problem awareness, search, answer, and action

Files in this skill

  • SKILL.md2.8 KB
  • assets/check-context.py14.6 KB
  • assets/context-contract.json9.6 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

View all in ai-agents →

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.