red-team-business

Adversarially audit a business, product, launch, pricing page, or project plan. Use when Kevin asks to red-team an idea, attack a business, find blind spots, run a premortem, simulate competitors/customers/investors, or pressure-test pricing/moat/distribution.

Category
Productivity
Skill
242 of 261
Source
Kevin wiki

Use this when the user wants the business or product attacked before the market does it for free.

Inputs

Gather only what is necessary:

  • product or project description
  • target customer
  • business model or pricing
  • distribution channel
  • current traction or constraints
  • what decision the audit should inform

If any of these are missing, proceed with explicit assumptions unless the missing fact changes the answer.

Seven Lenses

Run the audit through these lenses, adapted from the Claude Red Team Mode bookmark:

  1. Competitor War Game: how a ruthless competitor steals users, undercuts positioning, or makes the product irrelevant within 12 months.
  2. Customer Betrayal Sim: the honest cancellation email from a once-happy user.
  3. Investor Teardown: the three holes a skeptical investor would press on, plus the question the team cannot yet answer.
  4. Single Point Of Failure Scan: people, tool/platform, revenue, supply-chain, and channel dependencies.
  5. Market Shift Scenario: three plausible 18-month shifts and the early-warning signs.
  6. Pricing Pressure Test: why a price-sensitive buyer says no, the churn threshold, and the cheaper substitute.
  7. Attack Surface Report: top vulnerabilities ranked by likelihood and severity, plus the blind spot outside the builder's current frame.

Output Shape

Return:

  • Verdict: one paragraph naming the largest risk.
  • Critical vulnerabilities: ranked table with likelihood, severity, evidence, and mitigation.
  • Adversarial narratives: concise competitor/customer/investor versions.
  • Signals to watch: what would prove the risk is real.
  • Next moves: 3-5 actions that harden the plan.
  • Non-issues: fears that look dramatic but are not load-bearing.

Rules

  • Be concrete. Name the move, channel, substitute, failure mode, or missing evidence.
  • Do not use generic startup advice unless it directly follows from the user's context.
  • Treat model output as hypotheses. Recommend checks that can falsify the scary story.
  • If the project is early, optimize for learning speed. If it is live, optimize for risk reduction and retention.
  • Link durable insights back to wiki/concepts/red-team-your-business.md when updating the wiki.