dreamcontext 0.5.0

package/dist/skill-packs/business-idea-validation/stage-definitions.md

@@ -0,0 +1,658 @@
+---
+name: biv-stage-definitions
+description: >
+  Defines all 9 stages of the Business Idea Validation pipeline: inputs,
+  outputs, success criteria, kill signals, and confidence scoring rubric
+  for each stage. Referenced by the orchestrator and all specialist agents.
+---
+
+# Stage Definitions — Business Idea Validation Pipeline
+
+Each stage follows the same structure:
+- **Agent**: which specialist executes it
+- **Inputs**: which prior artifacts it consumes
+- **Deliverable**: what it writes to disk
+- **Success Criteria**: what PASS looks like
+- **Kill Signals**: specific red flags that trigger KILL
+- **Confidence Rubric**: how to score 1-5
+
+---
+
+## Stage 1: Problem Discovery
+
+**Agent:** `biv-researcher`
+**Inputs:** `00-brief.md` (user's idea brief)
+**Deliverable:** `01-problem-research.md`
+
+### What the agent does
+- Searches Reddit (via `site:reddit.com` advanced queries), forums, Quora
+  for real people expressing this problem
+- Mines 1-star and 2-star app store reviews of related products
+- Checks Google Trends for demand trajectory
+- Quantifies problem severity using evidence (upvotes, comment counts,
+  emotional language intensity)
+- Collects 5-15 real quotes as evidence
+
+### Deliverable structure
+```markdown
+## Problem Statement
+[One clear sentence defining the validated problem]
+
+## Evidence Summary
+- **Forums/Reddit**: [# of threads found, key subreddits, date range]
+- **Review Mining**: [which apps reviewed, common complaints]
+- **Demand Trajectory**: [Google Trends signal: growing/stable/declining]
+- **Search Volume Signals**: [relevant keywords and approximate demand]
+
+## Top Pain Points (ranked by severity)
+1. [Pain point] — Evidence: "[real quote]" (source, upvotes)
+2. ...
+3. ...
+
+## Existing Alternatives
+[How people currently solve this — tools, workarounds, manual processes]
+
+## Kill Signal Check
+- [ ] Problem is real and evidenced by multiple independent sources
+- [ ] Problem is current (threads/reviews from last 12 months)
+- [ ] People express willingness to pay or have tried paid alternatives
+- [ ] Google Trends shows stable or growing interest
+
+## Confidence: [1-5]
+## Kill Signals: [list, or "none"]
+## Executive Summary: [≤120 words]
+```
+
+### Success Criteria (PASS)
+- Found ≥5 independent sources confirming the problem
+- At least 3 pain points with real quotes
+- Google Trends not declining
+- People are already spending money on imperfect alternatives
+
+### Kill Signals
+- Google Trends shows sustained decline (>2 years downward)
+- Cannot find anyone expressing this problem online (< 3 sources)
+- Problem exists but is trivially solved by free tools already
+- Problem is purely hypothetical — no real-world evidence found
+
+### Confidence Rubric
+| Score | Criteria |
+|---|---|
+| 5 | 10+ sources, strong quotes, growing trend, paid alternatives exist |
+| 4 | 5-10 sources, good quotes, stable trend, some paid alternatives |
+| 3 | 3-5 sources, decent quotes, flat trend, free alternatives only |
+| 2 | 1-3 sources, weak quotes, unclear trend |
+| 1 | Cannot find evidence the problem exists at meaningful scale |
+
+---
+
+## Stage 2: Customer Analysis
+
+**Agent:** `biv-customer-analyst`
+**Inputs:** `00-brief.md`, `01-problem-research.md`
+**Deliverable:** `02-customer-analysis.md`
+
+### What the agent does
+- Identifies 2-4 distinct customer segments from the problem research
+- Scores each segment using the 3 Question Stud framework (Size, Pain:Payment, Accessibility)
+- Selects one beachhead segment
+- Creates 1-2 detailed personas for the beachhead
+- Defines buyer vs. user persona if they differ (B2B scenarios)
+
+### Deliverable structure
+```markdown
+## Customer Segments Identified
+| Segment | Size (1-10) | Pain:Payment (1-10) | Accessibility (1-10) | Total |
+|---|---|---|---|---|
+| [Segment A] | | | | |
+| [Segment B] | | | | |
+| ...         | | | | |
+
+## Beachhead Selection
+**Winner:** [Segment name]
+**Justification:** [Why this segment scores highest and is the right starting point]
+
+## Persona 1: [Name] (User Persona)
+- **Profile:** [Job title, age range, background, responsibilities]
+- **Goals & Motivations:** [What success looks like for them]
+- **Pains & Frustrations:** [Specific to the validated problem]
+- **Current Solutions:** [What tools/workarounds they use now]
+- **Quote:** "[A representative quote from research]"
+
+## Persona 2: [Name] (Buyer Persona — if different from user)
+[Same structure]
+
+## Confidence: [1-5]
+## Kill Signals: [list, or "none"]
+## Executive Summary: [≤120 words]
+```
+
+### Success Criteria (PASS)
+- At least 2 distinct segments identified with scoring
+- Clear beachhead selected with justified reasoning
+- At least 1 detailed persona with goals, pains, and current solutions
+- Persona is grounded in Stage 1 research, not hypothetical
+
+### Kill Signals
+- All segments score below 4/10 on Pain:Payment ratio
+- No segment is accessible (all score ≤ 2 on Accessibility)
+- The people who have the problem cannot pay for a solution
+
+### Confidence Rubric
+| Score | Criteria |
+|---|---|
+| 5 | 3+ segments scored, clear winner, personas grounded in rich data |
+| 4 | 2-3 segments, reasonable winner, personas well-defined |
+| 3 | 2 segments, scores close, personas partially hypothetical |
+| 2 | Only 1 vague segment, persona mostly assumed |
+| 1 | Cannot define who has this problem with any specificity |
+
+---
+
+## Stage 3: Market Sizing
+
+**Agent:** `biv-market-analyst`
+**Inputs:** `00-brief.md`, `01-problem-research.md`, `02-customer-analysis.md`
+**Deliverable:** `03-market-sizing.md`
+
+### What the agent does
+- Calculates TAM, SAM, SOM using both top-down and bottom-up methods
+- Sources data from industry reports, census data, LinkedIn, directories
+- Cross-validates the two methods for sanity
+- Applies the ambition-level gate (VC vs. bootstrapped)
+
+### Deliverable structure
+```markdown
+## Top-Down Analysis
+- **TAM**: $[X] — [source and methodology]
+- **SAM**: $[X] — [filters applied: geography, segment, product fit]
+- **SOM**: $[X] — [realistic capture rate and justification]
+
+## Bottom-Up Analysis
+- **Target customers**: [# in beachhead segment]
+- **Annual price assumption**: $[X]/year
+- **SOM**: [# customers] × $[price] = $[X]
+- **SAM**: [total segment size] × $[price] = $[X]
+- **TAM**: [global expansion] × $[price] = $[X]
+
+## Cross-Validation
+[Do top-down and bottom-up align? If not, which is more credible and why?]
+
+## Ambition Gate
+- **Ambition level**: [VC / Bootstrapped]
+- **VC threshold**: TAM > $1B, SAM supports $100M+ revenue → [PASS/FAIL]
+- **Bootstrapped threshold**: SOM shows path to $1-10M/year → [PASS/FAIL]
+
+## Confidence: [1-5]
+## Kill Signals: [list, or "none"]
+## Executive Summary: [≤120 words]
+```
+
+### Success Criteria (PASS)
+- Both top-down and bottom-up calculations completed
+- TAM/SAM/SOM documented with sources
+- Cross-validation shows reasonable alignment (within 3x)
+- Ambition gate passes for the user's stated level
+
+### Kill Signals
+- SOM < $500K for bootstrapped, or SAM < $500M for VC
+- Top-down and bottom-up diverge by >10x (one is likely wrong)
+- Cannot find any data to estimate market size (too niche)
+- Market is demonstrably shrinking (corroborates declining trends from Stage 1)
+
+### Confidence Rubric
+| Score | Criteria |
+|---|---|
+| 5 | Both methods align, strong data sources, clear ambition-gate pass |
+| 4 | Both methods completed, reasonable alignment, gate passes |
+| 3 | One method strong, one weak; gate barely passes |
+| 2 | Only one method possible, limited data, gate questionable |
+| 1 | Cannot size the market with any credibility |
+
+---
+
+## Stage 4: Competitive Intelligence
+
+**Agent:** `biv-researcher`
+**Inputs:** `00-brief.md`, `01-problem-research.md`, `02-customer-analysis.md`, `03-market-sizing.md`
+**Deliverable:** `04-competitive-intel.md`
+
+### What the agent does
+- Identifies direct, indirect, potential, and substitute competitors
+- Creates deep dossiers on the market leader + top 2-3 direct competitors
+- Builds a feature comparison matrix
+- Creates a 2x2 competitive positioning matrix
+- Identifies the unoccupied strategic position (the gap)
+
+### Deliverable structure
+```markdown
+## Competitor Map
+| Competitor | Type | Est. Revenue | Target Customer | Key Strength |
+|---|---|---|---|---|
+| [Name] | Direct | | | |
+| [Name] | Indirect | | | |
+| ...    | ...      | | | |
+
+## Deep Dossier: [Market Leader Name]
+- **Size & Revenue:** [estimates]
+- **User Base:** [who, how many]
+- **Brand & Messaging:** [core message, perception]
+- **Strengths:** [2-3 key strengths]
+- **Weaknesses:** [2-3 from reviews/forums]
+- **Innovation Speed:** [feature launches in last 1-2 years]
+
+## Deep Dossier: [Direct Competitor 1]
+[Same structure]
+
+## Feature Comparison Matrix
+| Feature | Our Concept | Leader | Comp. 1 | Comp. 2 |
+|---|---|---|---|---|
+| [Feature 1] | | | | |
+| ...         | | | | |
+
+## 2x2 Competitive Matrix
+- **Axis X:** [e.g., Price: Low → High]
+- **Axis Y:** [e.g., Complexity: Simple → Powerful]
+- **Positions:** [where each competitor sits]
+- **Empty quadrant:** [the opportunity]
+
+## Strategic Position
+"The market is crowded with [type A] and [type B], but no one is serving
+customers who need [our unique position]."
+
+## Confidence: [1-5]
+## Kill Signals: [list, or "none"]
+## Executive Summary: [≤120 words]
+```
+
+### Success Criteria (PASS)
+- At least 3 competitors identified and categorized
+- Deep dossier on market leader with strengths AND weaknesses
+- Feature matrix showing at least one differentiation opportunity
+- 2x2 matrix revealing an unoccupied or underserved quadrant
+
+### Kill Signals
+- Zero competitors found (almost always means no market, not an opportunity)
+- All competitors have > 4.5 star ratings and dominant market share with no complaints
+- The identified gap is trivially small or not valued by customers
+- A well-funded competitor launched the same concept in the last 6 months
+
+### Confidence Rubric
+| Score | Criteria |
+|---|---|
+| 5 | Rich competitor data, clear gap, verified complaints, defensible position |
+| 4 | Good data, reasonable gap, some complaints found |
+| 3 | Moderate data, gap exists but unclear if valued |
+| 2 | Limited competitor info, gap is vague |
+| 1 | No competitors OR no discernible gap |
+
+---
+
+## Stage 5: Customer Validation (AI-Simulated)
+
+**Agent:** `biv-customer-analyst`
+**Inputs:** All prior artifacts (stages 0-4)
+**Deliverable:** `05-customer-validation.md`
+
+### What the agent does
+- Synthesizes Reddit/forum data from Stage 1 with personas from Stage 2
+- Simulates customer responses to the proposed solution concept
+- Evaluates willingness-to-pay signals from real-world evidence
+- Tests the value proposition against the personas' stated pains
+- Identifies potential objections and adoption barriers
+- **Confidence is capped at 3** — synthetic validation cannot exceed moderate confidence
+
+### Deliverable structure
+```markdown
+> **Note:** This is AI-simulated validation based on forum/review data.
+> Real customer interviews would increase confidence. Treat findings as
+> hypotheses to verify, not proven facts.
+
+## Simulated Response: [Persona 1 Name]
+- **Reaction to core value prop:** [positive/neutral/negative + reasoning]
+- **Willingness to pay:** [evidence from forum data — existing spend, price sensitivity]
+- **Likely objections:** [based on complaints about similar products]
+- **Adoption barriers:** [switching costs, habits, trust]
+- **Would they try it?** [assessment + reasoning]
+
+## Simulated Response: [Persona 2 Name]
+[Same structure]
+
+## Cross-Persona Patterns
+- **Strongest signal:** [what resonates across personas]
+- **Biggest risk:** [what could prevent adoption]
+- **Price sensitivity:** [what the data suggests about willingness to pay]
+
+## Synthetic Validation Score
+- Problem resonance: [1-5]
+- Solution fit: [1-5]
+- Willingness to pay: [1-5]
+- Adoption likelihood: [1-5]
+
+## Confidence: [1-3] ← CAPPED (synthetic data)
+## Kill Signals: [list, or "none"]
+## Executive Summary: [≤120 words]
+```
+
+### Success Criteria (PASS)
+- At least one persona shows strong positive signals
+- Willingness-to-pay evidence exists (people already spending on alternatives)
+- No dominant objection that the proposed solution cannot address
+- Adoption barriers are surmountable
+
+### Kill Signals
+- Personas' stated needs directly contradict the proposed solution
+- All willingness-to-pay signals are negative (people expect free solutions)
+- Dominant objection that cannot be addressed (regulatory, trust, etc.)
+- The competitive alternative is "good enough" — switching cost exceeds benefit
+
+### Confidence Rubric (capped at 3)
+| Score | Criteria |
+|---|---|
+| 3 | Strong forum evidence, clear WTP signals, resonates with personas |
+| 2 | Moderate evidence, mixed signals, some alignment |
+| 1 | Weak or contradictory evidence, poor persona fit |
+
+---
+
+## Stage 6: Synthesis & Hypothesis
+
+**Agent:** `biv-strategist`
+**Inputs:** All prior artifacts (stages 0-5)
+**Deliverable:** `06-synthesis-hypothesis.md`
+
+### What the agent does
+- Creates the Opportunity Snapshot (one-page synthesis of all research)
+- Identifies the 2-3 riskiest assumptions
+- Formulates a testable solution hypothesis using the structured framework
+- Proposes 3-5 prioritized features using RICE scoring
+- Defines the "How Might We..." questions that bridge problem to solution
+
+### Deliverable structure
+```markdown
+## Opportunity Snapshot
+- **Core Problem:** [one sentence, from Stage 1]
+- **Target Persona:** [name, from Stage 2]
+- **Obtainable Market (SOM):** [from Stage 3]
+- **Key Competitor:** [from Stage 4]
+- **Competitive Gap:** [from Stage 4]
+- **Customer Signal:** [strongest evidence, from Stage 5]
+
+## Riskiest Assumptions
+1. **[Assumption]** — Impact: [1-10], Uncertainty: [1-10], Risk Score: [IxU]
+2. **[Assumption]** — Impact: [1-10], Uncertainty: [1-10], Risk Score: [IxU]
+3. **[Assumption]** — Impact: [1-10], Uncertainty: [1-10], Risk Score: [IxU]
+
+## Solution Hypothesis
+**We believe that** [target persona]
+**has a problem with** [core problem]
+**because** [root cause/insight].
+
+**If we build** [proposed solution — specific, simple description]
+**they will** [measurable action that proves we're right].
+
+## How Might We...
+1. HMW [reframe pain point 1 as opportunity]?
+2. HMW [reframe pain point 2 as opportunity]?
+3. HMW [reframe competitive gap as opportunity]?
+
+## Proposed Features (RICE scored)
+| Feature | Reach | Impact | Confidence | Effort | RICE Score |
+|---|---|---|---|---|---|
+| [Feature 1] | | | | | |
+| [Feature 2] | | | | | |
+| ...         | | | | | |
+
+## Confidence: [1-5]
+## Kill Signals: [list, or "none"]
+## Executive Summary: [≤120 words]
+```
+
+### Success Criteria (PASS)
+- Opportunity snapshot is coherent and internally consistent
+- At least 2 riskiest assumptions identified with scoring
+- Hypothesis is specific, testable, and has a measurable outcome
+- Features connect back to validated pain points
+
+### Kill Signals
+- Research from stages 1-5 contradicts itself (problem exists but market doesn't,
+  or customers want it but won't pay)
+- Cannot formulate a testable hypothesis — the solution is too vague
+- All riskiest assumptions have Risk Score > 80 (everything is uncertain)
+
+### Confidence Rubric
+| Score | Criteria |
+|---|---|
+| 5 | Snapshot coherent, hypothesis crisp, features strongly connected to research |
+| 4 | Good synthesis, clear hypothesis, reasonable feature list |
+| 3 | Some gaps in synthesis, hypothesis testable but broad |
+| 2 | Significant contradictions between stages, hypothesis weak |
+| 1 | Cannot synthesize a coherent story from the research |
+
+---
+
+## Stage 7: GTM Strategy
+
+**Agent:** `biv-strategist`
+**Inputs:** All prior artifacts (stages 0-6)
+**Deliverable:** `07-gtm-strategy.md`
+
+### What the agent does
+- Identifies where the target persona "lives" online (from Stage 1+2 data)
+- Evaluates 4-6 acquisition channels using ICE scoring
+- Selects 1-2 primary channels (the "wedge")
+- Crafts initial messaging using "Hook, Pain, Solution" framework
+- Outlines a plan to reach the first 100 users
+- Assesses organic marketability (TikTok/social virality potential)
+
+### Deliverable structure
+```markdown
+## Channel Analysis
+| Channel | Impact (1-10) | Confidence (1-10) | Ease (1-10) | ICE Score |
+|---|---|---|---|---|
+| [Channel 1] | | | | |
+| [Channel 2] | | | | |
+| ...         | | | | |
+
+## Selected Wedge Channels
+1. **Primary:** [channel] — [why it's the best fit]
+2. **Secondary:** [channel] — [backup / complementary]
+
+## Messaging Framework
+- **Hook:** [attention-grabber using customer language from Stage 1]
+- **Pain:** [agitate the problem]
+- **Solution:** [introduce the product]
+- **CTA:** [specific action: waitlist, trial, demo]
+
+## First 100 Users Plan
+1. [Week 1-2: specific action with expected result]
+2. [Week 3-4: specific action with expected result]
+3. [Month 2: specific action with expected result]
+
+## Marketability Assessment
+- **Social virality potential:** [high/medium/low + evidence]
+- **Content potential:** [can this topic generate organic content?]
+- **Controversy factor:** [does the topic naturally drive engagement?]
+
+## Confidence: [1-5]
+## Kill Signals: [list, or "none"]
+## Executive Summary: [≤120 words]
+```
+
+### Success Criteria (PASS)
+- At least 3 channels evaluated with ICE scoring
+- Clear primary channel selected with justified reasoning
+- Messaging framework uses actual customer language from Stage 1
+- First-100-users plan is specific and actionable
+
+### Kill Signals
+- No viable channel to reach the target persona (all score < 3 on ICE)
+- The persona is unreachable online (no forums, no social presence, no search intent)
+- Acquisition will require >$100 per user with no organic alternative
+
+### Confidence Rubric
+| Score | Criteria |
+|---|---|
+| 5 | Clear channel, strong messaging, organic virality confirmed |
+| 4 | Good channel, solid messaging, some organic potential |
+| 3 | Viable channel, messaging needs testing, unclear organic potential |
+| 2 | Channels are expensive/uncertain, messaging untested |
+| 1 | No viable path to reach the target customer |
+
+---
+
+## Stage 8: Financial Viability
+
+**Agent:** `biv-financial-analyst`
+**Inputs:** All prior artifacts (stages 0-7)
+**Deliverable:** `08-financial-viability.md`
+
+### What the agent does
+- Estimates Year 1 costs (one-time + monthly recurring)
+- Projects Year 1 revenue from SOM (Stage 3) and pricing assumptions
+- Calculates the Year 1 P&L
+- Computes unit economics: CAC, LTV, LTV:CAC ratio
+- Runs sensitivity analysis on key levers (price, churn, acquisition cost)
+- Checks if the model improves in Year 2
+
+### Deliverable structure
+```markdown
+## Year 1 Cost Estimate
+| Category | One-Time | Monthly | Annual |
+|---|---|---|---|
+| Development (founder time) | | | |
+| Design / freelancers | | | |
+| Infrastructure / hosting | | | |
+| Marketing / ads | | | |
+| Software / tools | | | |
+| Legal / admin | | | |
+| **Total** | **$X** | **$X/mo** | **$X** |
+
+## Year 1 Revenue Projection
+- **Target customers (Year 1):** [from SOM]
+- **Price per customer:** $[X]/year
+- **Projected revenue:** $[X]
+- **Year 1 P&L:** $[revenue - costs] = $[X]
+
+## Unit Economics
+- **CAC:** $[marketing spend / new customers]
+- **LTV:** $[monthly revenue per customer / monthly churn rate]
+- **LTV:CAC Ratio:** [X]:1
+- **Health assessment:** [healthy ≥3:1 / warning 1-3:1 / unsustainable <1:1]
+- **Payback period:** [months to recover CAC]
+
+## Sensitivity Analysis
+| Scenario | Change | Impact on LTV:CAC |
+|---|---|---|
+| Price +50% | $X → $Y | [ratio] |
+| Price -25% | $X → $Y | [ratio] |
+| Churn halved | X% → Y% | [ratio] |
+| CAC doubled | $X → $Y | [ratio] |
+
+## Year 2 Outlook
+- **Projected growth:** [customer count, revenue]
+- **Path to profitability:** [when does revenue exceed costs?]
+- **Key lever:** [what single change most improves the model?]
+
+## Confidence: [1-5]
+## Kill Signals: [list, or "none"]
+## Executive Summary: [≤120 words]
+```
+
+### Success Criteria (PASS)
+- Cost and revenue estimates documented with assumptions
+- LTV:CAC ≥ 3:1 (or a clear path to it within 12 months)
+- Sensitivity analysis shows the model isn't fragile (survives 1-2 adverse changes)
+- Year 2 shows meaningful improvement
+
+### Kill Signals
+- LTV:CAC < 1:1 even with optimistic assumptions
+- Year 1 costs exceed available funding/runway with no revenue path
+- Sensitivity analysis shows model breaks under any single adverse change
+- Unit economics require unrealistic churn rates (< 1%) to work
+
+### Confidence Rubric
+| Score | Criteria |
+|---|---|
+| 5 | Strong unit economics, robust under sensitivity, clear profitability path |
+| 4 | Good economics, some sensitivity risk, profitability within 18 months |
+| 3 | Marginal economics, sensitive to assumptions, profitability uncertain |
+| 2 | Weak economics, model fragile, requires significant optimism |
+| 1 | Math doesn't work under any reasonable scenario |
+
+---
+
+## Stage 9: Go/No-Go Decision
+
+**Agent:** `biv-decision-gate`
+**Inputs:** All prior artifacts (stages 0-8), `pipeline-state.json`
+**Deliverable:** `09-go-no-go-decision.md`
+
+### What the agent does
+- Reads all 8 stage artifacts and the pipeline state
+- Scores the idea across 8 dimensions (each 1-5)
+- Accounts for stages that passed with risk (confidence penalties)
+- Produces a total score and verdict
+- Writes a clear recommendation with specific next steps
+
+### Deliverable structure
+```markdown
+## Decision Scorecard
+
+### Problem-Solution Fit (Desirability)
+| Dimension | Score (1-5) | Evidence Source | Notes |
+|---|---|---|---|
+| Pain Level | | Stage 1 | |
+| Market Size | | Stage 3 | |
+| Competitive Edge | | Stage 4 | |
+| Customer Signal | | Stage 5 | |
+
+### Go-To-Market Fit (Feasibility)
+| Dimension | Score (1-5) | Evidence Source | Notes |
+|---|---|---|---|
+| Channel Clarity | | Stage 7 | |
+| Acquisition Cost | | Stage 8 | |
+
+### Business Model Fit (Viability)
+| Dimension | Score (1-5) | Evidence Source | Notes |
+|---|---|---|---|
+| Profitability Path | | Stage 8 | |
+| Founder Conviction | | Brief + overall | |
+
+### Total Score: [X] / 40
+
+## Verdict
+- **GO** (30-40): Strong evidence across all three pillars.
+- **PIVOT** (18-29): Positive signals exist but significant weaknesses in [areas].
+- **NO-GO** (8-17): Evidence does not support this idea in its current form.
+
+**Verdict: [GO / PIVOT / NO-GO]**
+
+## Pipeline Health
+- Stages passed clean: [list]
+- Stages passed with risk: [list + what the risk was]
+- Strongest dimension: [which + why]
+- Weakest dimension: [which + why]
+
+## Recommendation
+[2-3 paragraphs: what to do next, specific and actionable]
+
+## If PIVOT — Where to Re-enter
+[Which stage to re-run and what to change]
+
+## Executive Summary: [≤150 words]
+```
+
+### Kill Signals
+Stage 9 does not define its own kill signals. It aggregates evidence from
+stages 1-8. The verdict thresholds (NO-GO at 8-17) serve as the final gate.
+Any kill signals from prior stages that were overridden via "weak_proceed"
+are reflected as -1 scoring penalties, not as independent kill triggers here.
+
+### Scoring Notes
+- If Stage 5 (Customer Validation) was AI-simulated, cap its "Customer Signal"
+  dimension at 3 regardless of the simulated score
+- If any stage passed with risk ("weak_proceed"), apply a -1 penalty to its
+  corresponding dimension score
+- "Founder Conviction" is assessed from the user's original brief — how
+  specific, passionate, and knowledgeable they were about the domain