precoil-emt

A free Claude skill for testing business ideas using the Extract → Map → Test system by David J. Bland

Instalación
CLI
npx skills add https://github.com/david-precoil/testing-business-ideas-with-claude --skill precoil-emt

Instala esta habilidad con la CLI y comienza a usar el flujo de trabajo SKILL.md en tu espacio de trabajo.

Última actualización 4/22/2026

Testing Business Ideas with Claude

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.


Example EMT Output (Condensed)

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."

Extract

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.

Map

Highest‑Risk Assumption
AI predictions align with health concerns cat owners already worry
about.

Risk Type
Desirability

Test

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.


What the Skill Does

The skill runs a guided sequence:

  1. Extract -- surfaces assumptions across Desirability, Viability,
    and Feasibility\
  2. Map -- identifies the highest‑risk assumptions to test\
  3. Test -- generates an experiment brief to reduce uncertainty

Output typically includes:

  • a structured assumption map\
  • prioritization of the biggest risks\
  • a clear experiment brief

Who This Is For

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:

  • product leaders evaluating new initiatives
  • innovation and strategy teams pressure‑testing concepts
  • corporate venture or incubation teams
  • consultants facilitating assumption mapping
  • organizations introducing structured experimentation

Used in Practice

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...


What it does not include

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


How to install

  1. Download testing-business-ideas-with-claude.zip
  2. Open Claude.ai and go to Settings → Capabilities
  3. Ensure Code execution and file creation is enabled
  4. Go to Customize → Skills
  5. Click "+""Upload a skill" and upload the ZIP file

Requires a free, Pro, Max, Team, or Enterprise Claude account.


How to use

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:

  • "Pressure test this idea"
  • "What are the riskiest assumptions here?"
  • "Extract assumptions from this strategy"
  • "Identify hidden risk in this initiative"

Note for artifact and app builders

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.


Read before installing

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.


License

This skill is released under the MIT license. The Precoil Experiment Library is a separate licensed asset and is not included here.


About Precoil

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.

Precoil.com · Experiment Library