agentsclimarketplace

Business design

Skill idimsh/tdds-business-skills/business-design

Portable SKILL.md agent skills for Claude, Codex, and other AI coding agents — audits, legal, design, and prompt engineering

Install
npx -y skills add idimsh/tdds-business-skills --skill business-design

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use when the user wants to design a business model, validate a venture idea, or create a structured business plan. Triggered by requests like "think through a business idea", "business model", "validation plan", "Business Design", "Geschäftsidee durchdenken", "Business Model erstellen", or when a venture needs a structured thesis, Lean Canvas, hypotheses, roadmap, and decision matrix.

SKILL.md

8.9 KB, as published. Nobody here has run it

Business Design Workshop

Interactive workshop that turns a venture idea into a structured, testable business design. Output: a design document with venture thesis, Lean Canvas, validation hypotheses, validation roadmap, and decision matrix.

Communicate in the user's language. If the project or user is German-language, use German in conversation and in the generated business design document. Be analytical, direct, and pragmatic.

Project Context

$ARGUMENTS

Agent Portability

This skill must work across Codex, Claude, and other SKILL.md-compatible agents.

  • Do not assume a specific slash-command system.
  • Do not rely on any agent-specific question tool. If a structured question tool is available, use it; otherwise ask one concise plain-text question.
  • Check common project instruction files when present: AGENTS.md, CLAUDE.md, README.md, .cursor/rules, .windsurfrules, and equivalent project docs.
  • Do not assume a specific file structure; adapt output paths to the repository. Prefer docs/plans/ only when it exists or matches project conventions.

Workflow: 5 Sections

Guide the user through one section at a time. Present each section, wait for approval, then continue. Ask one question per message, with multiple choice when useful.

digraph business_design {
  "Understand context" [shape=box];
  "Clarifying questions" [shape=box];
  "Venture Thesis" [shape=box];
  "Lean Canvas" [shape=box];
  "Hypotheses & Metrics" [shape=box];
  "Validation Roadmap" [shape=box];
  "Decision Matrix" [shape=box];
  "Write Design Doc" [shape=doublecircle];

  "Understand context" -> "Clarifying questions";
  "Clarifying questions" -> "Venture Thesis";
  "Venture Thesis" -> "Lean Canvas";
  "Lean Canvas" -> "Hypotheses & Metrics";
  "Hypotheses & Metrics" -> "Validation Roadmap";
  "Validation Roadmap" -> "Decision Matrix";
  "Decision Matrix" -> "Write Design Doc";
}

Step 0: Understand Context

Before asking questions, read existing docs if present:

  • docs/plans/ or equivalent planning docs: existing design docs and due-diligence outputs
  • Project instruction files: AGENTS.md, CLAUDE.md, README.md, .cursor/rules, .windsurfrules
  • Recent git commits: current project state

Then ask clarifying questions, one per message:

  1. Vision: Is this a one-off test or a long-term business?
  2. Revenue model: How should it make money, or is this still unclear?
  3. Target output: What should exist at the end?
  4. Document audience: Who is this document for: internal team, investors, sponsors, or another audience?

Adapt depth and tone to the answers. If revenue is unclear, do not force revenue assumptions. If the audience is investors, use a more evidence-oriented and persuasive tone with market data where available.


Section 1: Venture Thesis

Based on the answers, draft:

  • One-liner: One sentence describing the entire venture.
  • Core thesis: 2-3 sentences describing the market problem, solution, and connection between target groups.
  • Unfair advantage: 2-4 bullet points explaining what this venture has that others cannot easily copy.

Present the section and wait for approval before continuing.


Section 2: Lean Canvas

Create a Lean Canvas table with these fields:

FieldDescription
ProblemWhat is the core problem? Why does it exist?
Target groupsPrimary and secondary groups, if present, such as users and B2B partners.
Value propositionConcrete value for each target group.
SolutionHow the problem is solved. Be concrete, not abstract.
ChannelsHow the target groups will be reached.
Revenue streamsHow money is made. Use "TBD after validation" when honestly unclear.
Cost structureMain cost drivers.
Key metricsNumbers that show whether it works.
Unfair advantageWhy competitors cannot simply copy it.

Use Lean Canvas instead of the classic Business Model Canvas because it fits early-stage validation better. Fields may be TBD when assumptions are genuinely unknown.

Present the section and wait for approval.


Section 3: Hypotheses & Experiment Metrics

Create 2-5 testable hypotheses as a table:

#HypothesisPass criterionFail criterionMeasurement method
H1Clear, falsifiable statementConcrete thresholdConcrete thresholdExact measurement method

Rules for good hypotheses:

  • Each hypothesis must be falsifiable with clear pass/fail thresholds.
  • Measurement methods must be concrete, such as a database query, survey question, or analytics event.
  • Separate pre-launch hypotheses, testable through landing pages or ads, from post-launch hypotheses, testable only after product or event usage.
  • Sort by test order: what can be tested first?

Present the section and wait for approval.


Section 4: Validation Roadmap

Create a 3-phase roadmap.

Phase 1: Signal

  • Goal: Prove demand with minimal investment.
  • Concrete action plan as a table: step, action, timeline.
  • Estimated cost for Phase 1.
  • Which hypotheses are tested in Phase 1?
  • Stealth vs. open launch: Recommend stealth or cold traffic when the idea is unvalidated, because a personal network can create bias.

Phase 2: Repeat

  • Goal: Prove repeatability.
  • 4-5 bullet points describing what happens if Phase 1 succeeds.

Phase 3: Scale

  • Goal: Define a scalable model.
  • 4-5 bullet points covering expansion, revenue validation, and partnerships.

Present the section and wait for approval. Adapt if the user rejects a stealth approach.


Section 5: Decision Matrix

Create a decision matrix covering all realistic scenarios:

ScenarioHypothesis AHypothesis BDecision
Best casePassPassGo. Continue to the next phase.
Partial success APassFailGo with reservation. State what must change.
Partial success BFailPassPivot. State the exact pivot.
Wrong audiencePass with wrong audienceAnyPivot audience.
Worst caseFailFailKill or fundamental pivot. Run a post-mortem.

Add mandatory kill criteria so the team does not waste money:

  • Concrete threshold plus timeframe that triggers a stop.
  • Early warning signal that triggers immediate messaging changes.

Present the section and wait for approval.


After The Workshop

Save The Design Doc

Summarize all 5 sections in a document and save it using the repository's planning-document convention. If no convention exists, use:

docs/plans/YYYY-MM-DD-<project-name>-business-design.md

Format:

# Business Design: [Project Name]

**Date:** YYYY-MM-DD
**Status:** Approved
**Venture:** [Venture Name]
**Audience:** [Document audience]

---

## 1. Venture Thesis
[...]

## 2. Lean Canvas
[...]

## 3. Hypotheses & Experiment Metrics
[...]

## 4. Validation Roadmap
[...]

## 5. Decision Matrix
[...]

---

## Validation Tool
[Reference to landing page, MVP, or experiment implementing Phase 1]

Commit

Only commit if the user asked for commits or the repository workflow clearly expects it:

git add docs/plans/YYYY-MM-DD-*-business-design.md
git commit -m "Add [ProjectName] business design document"

Next Step

Ask the user what to do next:

The business design is ready. How do you want to continue?

Options:

  • Build validation tool: design or implement the landing page, MVP, or experiment.
  • Deepen due diligence: research market, competitors, assumptions, and risks.
  • Legal compliance check: if the skill legal-audit is available, load it and audit the project according to its rules. Otherwise, note that legal compliance (Impressum, Datenschutz, AGB, cookie consent) should be reviewed separately before launch.
  • Stop for now: no further action.

Rules

  1. One section at a time; never jump ahead.
  2. Require approval for each section before continuing.
  3. Ask one question per message; prefer multiple choice when useful.
  4. Be honest: TBD is better than invented assumptions.
  5. Hypotheses must be falsifiable; avoid soft wish statements.
  6. Kill criteria are mandatory.
  7. Use Lean Canvas instead of classic BMC for pre-validation ventures.
  8. Recommend cold-signal validation when the idea is unvalidated, because personal-network feedback is biased.
  9. Read existing docs before drafting final outputs.
  10. Conversation and document language should follow the user/project language; code, filenames, and commits should remain English unless the repository convention differs.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.