A free Claude skill for testing business ideas using the Extract → Map → Test system by David J. Bland
npx skills add https://github.com/david-precoil/testing-business-ideas-with-claude --skill precoil-emtInstallez cette compétence avec la CLI et commencez à utiliser le flux de travail SKILL.md dans votre espace de travail.
Run the Extract → Map → Test (EMT) system inside Claude to surface
the riskiest assumptions behind a new initiative, product concept, or
strategic bet before committing budget, teams, or development
resources.
Created by David J. Bland
Co‑author of Testing Business Ideas and creator of the Precoil EMT
System.
⭐ If this skill is useful, please star the repo so others can find it.
Idea being tested
"Add an AI feature that predicts health issues for cats to our wearable
cat health monitoring device for $15/month, starting with existing
customers in San Francisco."
Desirability - Cat owners want predictive insight into their pet's
health. - AI predictions would feel meaningfully better than existing
alerts.
Viability - Enough customers will pay $15/month to justify the
investment. - The pilot cohort is large enough to generate meaningful
conversion data.
Feasibility - Existing sensor data is sufficient to train a reliable
prediction model. - Predictions can be validated quickly enough to build
trust.
Highest‑Risk Assumption
AI predictions align with health concerns cat owners already worry
about.
Risk Type
Desirability
Experiment: Customer Interview
Goal: Validate whether predicted conditions match real owner concerns.
Pass signal - 70%+ of predicted conditions are named unprompted or
rated highly relevant.
The skill runs a guided sequence:
Output typically includes:
While EMT originated in startup experimentation practices, it is widely
used by product, innovation, and strategy teams inside established
organizations to evaluate initiatives before committing significant
resources.
Useful for:
The EMT system behind this skill has been facilitated in 1,000+
experimentation sessions with teams inside established organizations.
Organizations where the approach has been applied include teams at:
Adobe · Toyota · HP · DuPont and more...
The Precoil Experiment Library is not included in this skill.
The library contains 50+ experiment designs mapped to assumption
types and helps teams choose the fastest experiment to produce
evidence.
Explore the library:
https://www.precoil.com/library
testing-business-ideas-with-claude.zipRequires a free, Pro, Max, Team, or Enterprise Claude account.
Once installed, start a new conversation and prompt:
Run EMT on this idea:
[describe your product, strategy, or initiative]
Or use any of these trigger phrases:
If you're using this skill to build a Claude-powered artifact or browser-based app, note that image analysis in the Map phase works in claude.ai chat and Claude Code, but not in browser-based artifacts that call the Anthropic API directly. This is a CORS restriction — the API doesn't accept image payloads from browser fetch calls.
The full SKILL.md is readable above. It contains all instructions Claude follows when the skill is active — no hidden logic, no external calls, no data collection.
This skill is released under the MIT license. The Precoil Experiment Library is a separate licensed asset and is not included here.
Precoil helps organizations derisk major initiatives through structured business experimentation. The EMT system has been facilitated over 1,000 times across enterprise teams including Adobe, Toyota, HP, and DuPont.