ventureos 1.1.3 → 1.1.4

package/agents/pitch-master.md

@@ -51,24 +51,45 @@ These are your operating instructions for this VentureOS session. You are Claude
 
   <prompts>
     <prompt id="build-final-pitch">
-      Build the full 12-slide investor pitch deck.
+      Build the full 12-slide investor pitch deck using a narrative-first format.
+
       1. Load ALL completed artifacts from venture-state.yaml
       2. Read each artifact to extract the key evidence for each slide
-      3. Using the template at {project-root}/ventureOS/templates/pitch-deck.md, build all 12 slides:
-         Slide 1: Title + Hook (the undeniable market shift or opportunity)
-         Slide 2: Problem (validated customer pain — FIP data, quotes, journey)
-         Slide 3: Solution (wedge definition + vision for scaling)
-         Slide 4: Market Opportunity (TAM/SAM/SOM — top-down + bottom-up)
-         Slide 5: Product / Wedge Demo (prototype screenshots or description)
-         Slide 6: Business Model + Revenue Streams
-         Slide 7: Traction / Validation Evidence (desirability scores, pilots, conversions, A/B results)
-         Slide 8: Go-to-Market Strategy (channels, CAC hypothesis, sales process)
-         Slide 9: Competitive Landscape (differentiation map, wedge defensibility)
-         Slide 10: Financial Projections (3-5yr, unit economics, key milestones)
-         Slide 11: Team + Parent Organization Advantage + Ask
-         Slide 12: Appendix (risk assessment, detailed financials, experiment results)
-      4. For each slide: write compelling narrative content (not bullet dumps — tell the story)
-      5. Identify where Excalidraw visuals would strengthen the slide (charts, diagrams, maps)
+
+      3. For EVERY slide, use this exact structure — no exceptions:
+
+         ---
+         ### Slide N — [Slide Title]
+
+         **Headline:** [One punchy sentence the investor reads and remembers — max 12 words]
+
+         **Visual:** [What goes on screen: one dominant image, chart, stat, or quote — described concisely for a designer]
+
+         **You say:** [2–3 sentences the presenter delivers out loud — the story, not a list of facts]
+
+         **Proof:** [The one data point, quote, or evidence that makes this credible]
+         ---
+
+      4. Build all 12 slides in this sequence:
+         Slide 1:  Hook — the undeniable market shift or truth that makes this inevitable
+         Slide 2:  Problem — validated pain with FIP scores and a customer quote that lands hard
+         Slide 3:  Customer — the ICP's world: what their day looks like, what they fear, what they want
+         Slide 4:  Solution — the wedge entry point and the big vision arc
+         Slide 5:  Why Now — what changed in the last 2–3 years that opens this window
+         Slide 6:  Market — TAM/SAM/SOM with bottom-up reasoning, not just top-down numbers
+         Slide 7:  Product / Demo — the earliest testable version: what it does, what it replaces
+         Slide 8:  Business Model — how money flows, pricing, unit economics in one clear frame
+         Slide 9:  Traction / Validation — real or pilot evidence, experiment results, desirability scores
+         Slide 10: Go-to-Market — the first 100 customers: who, where, how, in what sequence
+         Slide 11: Team + Advantage — team credentials + parent org advantage + the ask
+         Slide 12: Appendix — risk assessment, detailed financials, experiment plan
+
+      5. Narrative arc rule:
+         - Slides 1–3: build the villain (the problem world)
+         - Slides 4–5: introduce the hero (the solution and timing)
+         - Slides 6–10: prove the business
+         - Slide 11: close the deal
+
       6. Save Markdown pitch to {output_folder}/{venture_name}/pitch/pitch-deck.md
     </prompt>
     <prompt id="build-checkin-pitch">

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "ventureos",
-  "version": "1.1.3",
+  "version": "1.1.4",
   "description": "VentureOS — AI-powered venture building framework. From raw idea to investor-ready pitch in 12 structured weeks.",
   "type": "module",
   "bin": {

package/workflows/3-find-pain/customer-pain-discovery/step-2-interviews.md

@@ -3,7 +3,7 @@
 <!-- Agent: Customer Discovery Specialist -->
 
 ## Objective
-Gather raw interview data using the selected mode: Toolkit (real interviews), Simulation (AI-generated), or Integration (external platform import). Process runs iteratively — run Step 3 (Synthesis) after each batch of 5-10 interviews.
+Gather raw interview data using the selected mode: Toolkit (real interviews), Simulation (AI-generated), or Integration (external platform import). Process runs iteratively — run Step 3 (Synthesis) after each batch of 10 interviews. Minimum 30 interviews total (25 customer + 5 expert) before proceeding to synthesis.
 
 ---
 
@@ -37,8 +37,8 @@ After every 5-10 interviews: prompt user to proceed to Step 3 (Synthesis) before
 ### 2S.1 Generate Customer Personas
 
 Using the ICP hypothesis and domain research:
-1. Generate 4-6 distinct customer personas representing the ICP
-2. Vary across: role, company size, industry vertical, level of pain, context
+1. Generate 25 distinct customer personas representing the ICP
+2. Vary across: role, company size, industry vertical, level of pain, context, geography, seniority, and tech-savviness
 3. Each persona: name, role, company context, day-in-the-life, technology stack, reporting structure
 
 ### 2S.2 Simulate Pain-Point Interviews
@@ -54,7 +54,7 @@ Save: `{output_folder}/{venture_name}/interviews/synthetic/synthetic-interviews.
 
 ### 2S.3 Simulate Expert Interviews
 
-Generate 2 domain expert personas (consultant, industry analyst, ex-practitioner):
+Generate 5 domain expert personas (consultant, industry analyst, ex-practitioner, market regulator or policy expert, academic researcher):
 1. Run expert interview script
 2. Simulate responses about: pain prevalence across the market, what companies typically do, where the biggest gaps are
 3. Label all output: "⚠️ SIMULATED EXPERT — AI-generated"

package/workflows/3-find-pain/customer-pain-discovery/step-5-pain-journey-map.md

@@ -115,7 +115,7 @@ Before declaring Phase 3 complete, verify all guiding questions:
 | What do people need? | Yes/No | |
 | Can the team prioritize the most important pain? | Yes/No | pain-atomization.md |
 | Is the pain frequent, intense, prevalent enough? | Yes/No | pain-atomization.md (FIP ≥ 3.0) |
-| Has the team validated the pain exists? (10+ interviews) | Yes/No | interview-synthesis.md |
+| Has the team validated the pain exists? (30+ interviews) | Yes/No | interview-synthesis.md |
 | Can they identify their ideal customer? | Yes/No | icp-profile.md |
 | Does the team understand who the buyer is? | Yes/No | icp-profile.md |
 
@@ -133,9 +133,9 @@ When Phase 3 is complete, brief the Venture Master and Product Strategist:
 5. What must the solution do to address the validated pain?
 
 **Phase 4 can begin when:**
-- Primary pain FIP ≥ 3.0 (weak signal — need more interviews)
+- Primary pain FIP ≥ 3.0 (weak signal — need more interviews, target 30+)
 - ICP is defined
-- At least 10 interviews completed
+- At least 30 interviews completed
 - Pain journey map shows at least one viable solution entry point
 
 ---

package/workflows/3-find-pain/customer-pain-discovery/workflow.md

@@ -63,7 +63,7 @@ Phase 3 is complete when ALL of the following guiding questions are answered wit
 - [ ] What do people need? (primary pain identified)
 - [ ] Can the team prioritize the most important pain to solve?
 - [ ] Is the pain frequent, intense, and/or prevalent enough to support a business? (FIP ≥ 3.0 average)
-- [ ] Has the team validated the pain exists? (minimum 10 interviews recommended)
+- [ ] Has the team validated the pain exists? (minimum 30 interviews recommended)
 - [ ] Can they identify their ideal customer? (ICP defined)
 - [ ] Does the team understand who the buyer is? (buyer vs. user mapped)
 

package/workflows/4-define-solution/wedge-design/step-2-value-propositions.md

@@ -40,7 +40,7 @@ Save cards to `{output_folder}/{venture_name}/concept-cards/card-{N}.md`
 
 **Option A: Real Concept Card Interviews**
 Using the interview script from Phase 3 (concept card section):
-1. Run 5-10 concept card interviews with ICP
+1. Run 15-20 concept card interviews with ICP
 2. Show each concept card and probe: resonance, differentiation, willingness to use/pay
 3. Capture scores and quotes per card
 

package/workflows/4-define-solution/wedge-design/step-4-solution-testing.md

@@ -11,7 +11,7 @@ Test the prototype with target customers. Gather usability and desirability feed
 
 Using the solution testing section of `interview-script.md`:
 
-**Conduct 5-10 solution testing interviews:**
+**Conduct 15-20 solution testing interviews:**
 1. Walkthrough the prototype with the ICP persona
 2. Ask participants to think aloud as they interact
 3. Probe: What's confusing? What's delightful? What would you do next?
@@ -22,7 +22,7 @@ Using the solution testing section of `interview-script.md`:
 **Synthetic Solution Testing (if needed):**
 Using "synthetic-solution-testing" tool from `synthetic-tools.csv`:
 1. Generate ICP personas from interview data
-2. Simulate 5+ personas experiencing the prototype walkthrough
+2. Simulate 15+ personas experiencing the prototype walkthrough
 3. Label ALL output: "⚠️ SIMULATED — Not real user testing"
 
 ---

package/workflows/autopilot/autopilot.md

@@ -132,12 +132,12 @@ Load and follow `{project-root}/ventureOS/agents/customer-discovery.md` — acti
 **Clara's task:**
 
 **Part A — Persona generation**
-Generate 6 distinct customer personas based on domain scan and market mapping findings.
-Vary across: role, company size, industry sub-vertical, geography, pain intensity.
+Generate 25 distinct customer personas based on domain scan and market mapping findings.
+Vary across: role, company size, industry sub-vertical, geography, pain intensity, seniority, and tech-savviness.
 For each persona define: name, role, company context, daily workflow, current tools, suspected pain level.
 
 **Part B — Synthetic interviews**
-Simulate a 30-minute pain interview for each of the 6 personas.
+Simulate a 30-minute pain interview for each of the 25 personas.
 Each interview must include:
 - Natural conversation (not Q&A format)
 - Genuine moments of resistance and surprise
@@ -145,9 +145,11 @@ Each interview must include:
 - Preliminary FIP signals embedded in responses
 
 **Part C — Expert interviews**
-Simulate 2 domain expert interviews (e.g. industry analyst, veteran practitioner).
+Simulate 5 domain expert interviews (e.g. industry analyst, veteran practitioner, former competitor employee, market regulator, academic researcher).
 Focus: pain prevalence across the market, why current solutions fail, what a breakthrough would look like.
 
+**Total: minimum 30 interviews (25 customer + 5 expert).**
+
 **All output labeled:**
 ```
 ⚠️ SIMULATED INTERVIEW — AI-generated persona. Not real customer data.
@@ -162,7 +164,7 @@ Update progress tracker: `[✓] Phase 2 — Customer Pain Discovery`
 ### Step 6 — Pain synthesis and FIP scoring (Clara)
 
 **Clara's task:**
-- Extract recurring pain themes across all 6 interviews
+- Extract recurring pain themes across all 30 interviews
 - Atomise into 4–6 distinct pain units
 - Score each on FIP framework using `{project-root}/ventureOS/scoring/pain-scoring.yaml`
 - Select primary pain: highest FIP AND most aligned with domain opportunity
@@ -359,17 +361,36 @@ Load and follow `{project-root}/ventureOS/agents/pitch-master.md` — activate P
 Run autonomously.
 
 **Petra's task:**
-Build a complete investor pitch deck covering:
-1. **Problem** — the primary pain with FIP evidence
-2. **Customer** — ICP profile with key signals from synthetic interviews
-3. **Solution** — the wedge and big vision
-4. **Why now** — market timing from domain scan
-5. **Market** — TAM / SAM / SOM with methodology
-6. **Business model** — revenue model and unit economics
-7. **Go-to-market** — first 100 customers strategy
-8. **Traction** (synthetic) — what the experiments are designed to prove
-9. **Team** — placeholder (to be filled by user)
-10. **The ask** — what the venture needs to reach the next milestone
+Build a complete investor pitch deck. Use the narrative-first format for every slide — no bullet dumps.
+
+**Each slide must follow this exact structure:**
+```
+---
+### Slide N — [Slide Title]
+
+**Headline:** [One punchy sentence the investor reads and remembers — max 12 words]
+
+**Visual:** [What goes on screen: one dominant image, chart, stat, or quote — described concisely]
+
+**You say:** [2–3 sentences the presenter delivers out loud — the story, not a list]
+
+**Proof:** [The one data point, quote, or evidence that makes this real]
+---
+```
+
+**Slides to build (10):**
+1. **Hook** — the market shift or undeniable truth that makes this venture inevitable
+2. **Problem** — the primary pain with FIP evidence and a customer quote that lands like a gut-punch
+3. **Customer** — ICP profile, their world before the solution, what they fear and want
+4. **Solution** — the wedge and the big vision arc (smallest entry → where this goes)
+5. **Why now** — the timing argument: what changed in the last 2–3 years that opens this window
+6. **Market** — TAM / SAM / SOM with bottom-up reasoning, not just numbers
+7. **Business model** — how money flows, pricing, unit economics in one clear visual
+8. **Go-to-market** — the first 100 customers: who, where, how, in what sequence
+9. **Traction** (synthetic) — what the 90-day experiments are designed to prove, and what GO looks like
+10. **Team + The ask** — placeholder team section + what the venture needs to reach the next milestone
+
+**Narrative arc rule:** Slides 1–3 build the villain (the problem). Slides 4–5 introduce the hero (the solution and timing). Slides 6–8 prove the business. Slides 9–10 close the deal.
 
 **Output:** Save as `{output_folder}/{venture_name}/autopilot/pitch-deck.md`