8 AI business advisors that stress-test your startup ideas using real frameworks from ESADE's MIE program. Built as a skill for Claude Code, Codex, and similar AI coding tools.
Hey MIE folks — if you have a business idea for class, run it through this before your pitch. It will find the holes your professors will find. Use it, fork it, improve it.
git clone https://github.com/b1rd33/mie-advisor.git
cd mie-advisor
claudeThe skill loads automatically. Type /mie-advisor and go.
git clone https://github.com/b1rd33/mie-advisor.git
cd mie-advisor
codex exec -m gpt-5.4 --full-auto 'Run /mie-advisor challenge "my assumption here"'Clone the repo and open it in your tool. If it reads .claude/skills/ markdown files, it will pick up the skill automatically. If not, copy the content of .claude/skills/mie-advisor.md into your system prompt or custom instructions.
You give it a business idea. It runs 8 expert advisors — each with a different discipline and real course frameworks — then they debate each other and score the result.
/mie-advisor analyze "An app that helps freelancers optimize tax deductions"
Full analysis: 8 advisors + debate + scoring + action plan. Takes a few minutes.
/mie-advisor challenge "Restaurants will pay $99/month for this"
Quick stress-test: 3 advisors attack your assumption. Under 2 minutes.
/mie-advisor consult strategy "Is there a real moat here?"
Single advisor deep dive. Pick the lens you want.
| # | Name | Discipline | What they look for |
|---|---|---|---|
| 01 | Lena Voss | Design Thinking | User empathy, Jobs To Be Done, behavioral assumptions |
| 02 | Adrian Vale | Product Management | MVP definition, prioritization, product-market fit |
| 03 | Helena Mora | Innovation Management | Innovation type, S-curves, disruption patterns |
| 04 | Kevin Coyne | Advanced Strategy | Competitive advantage, moats, wedge analysis |
| 05 | Maya Chen | Entrepreneurship | Validation, unit economics, lean experiments |
| 06 | Nico Ferran | Exploring Opportunity | Market sizing, beachhead, GTM, pricing |
| 07 | Nadia Soler | Business in Society | Stakeholder mapping, externalities, ESG |
| 08 | Cole Mercer | Silicon Valley Insights | VC lens, kill price math, distribution, scaling |
Each advisor uses real named frameworks from ESADE courses — not generic advice. They cite specific models like the Three-Leg Test, Beer Zone pricing, Empathy Maps, and Customer Development.
Phase 1 All 8 advisors analyze your idea independently
Phase 2 Structured debate — odd-numbered advisors challenge even, defenders respond
Phase 3 Harsh scoring on 8 dimensions (0-100, anti-inflation rules apply)
Phase 4 Final report with executive summary, verdicts, and action plan
The debate surfaces risks no single perspective catches. In testing, it found things like "the JUICER/Wendy's pricing backlash means your category has a trust problem" — insights that only emerged from advisors arguing with each other.
A /mie-advisor analyze run produces:
- Executive summary with go/no-go recommendation
- 8 individual assessments with verdicts (green/yellow/red)
- Debate synthesis: consensus, unresolved disagreements, new insights
- Score breakdown across 8 dimensions
- Action plan: this week / this month / this quarter
- Top 10 critical questions ranked by importance
design-thinking · product-mgmt · innovation · strategy · entrepreneurship · opportunity · society · silicon-valley
When new courses are completed:
- Add a knowledge file in
auto-advisor/knowledge/[course-name].md - Add a persona file in
auto-advisor/.claude/agents/[NN]-[name]-advisor.md - Update the advisor table in
.claude/skills/mie-advisor.md
Planned:
- Finance & Accounting
- Marketing & Customer Acquisition
- Operations & Supply Chain
- Legal & IP Strategy
- Claude Code or Codex CLI or any AI tool that reads skill files
- The advisors use web search for real market data — works best with internet access enabled
Licensed for ESADE MIE students and alumni only. See LICENSE for details.