convoke-agents 2.0.0

package/_bmad/bme/_vortex/workflows/proof-of-value/steps/step-04-test.md

@@ -0,0 +1,107 @@
+---
+step: 4
+workflow: proof-of-value
+title: Run the Validation Test
+---
+
+# Step 4: Run the Validation Test
+
+You have designed the experiment. Now execute it with discipline — capturing real market signals without contaminating the results.
+
+## Why This Matters
+
+Experiment execution is where rigor meets reality. A well-designed validation test can be ruined by sloppy execution: changing the price mid-test, adding participants who do not match your criteria, interpreting early results and adjusting course prematurely. The goal is clean data from a well-controlled experiment. Every deviation from the protocol weakens the signal. Run the experiment exactly as designed, document everything, and resist the urge to intervene until you have enough data to draw a conclusion.
+
+## Your Task
+
+### 1. Pre-Launch Checklist
+
+Before launching the test, confirm each element is in place:
+
+- [ ] **Stimulus is live** — Landing page, pricing page, prototype, or concierge service is ready and accessible
+- [ ] **Measurement is active** — Analytics, payment tracking, or manual logging is capturing data
+- [ ] **Recruitment is ready** — Ads, outreach sequences, or partner introductions are queued
+- [ ] **Success criteria are documented** — Pre-committed thresholds from Step 2 and Step 3 are written down and visible
+- [ ] **Abort criteria are clear** — You know under what conditions you stop early
+- [ ] **Team is briefed** — Anyone involved knows the protocol and understands they must not deviate
+
+### 2. Execute the Test
+
+Track your experiment in real time using this log:
+
+**Experiment Execution Log:**
+
+| Date | Participants Reached | Conversions | Conversion Rate | Notes |
+|------|---------------------|-------------|-----------------|-------|
+| Day 1 | | | | |
+| Day 2 | | | | |
+| Day 3 | | | | |
+| ... | | | | |
+| **Total** | | | | |
+
+**Behavioral Signals to Capture:**
+
+| Signal Type | What to Record |
+|-------------|---------------|
+| **Payment signals** | Deposits placed, credit cards entered, invoices requested, pre-orders completed |
+| **Time investment signals** | Demo requests, onboarding time, configuration effort, follow-up meetings scheduled |
+| **Switching signals** | Competitors mentioned, comparison shopping behavior, cancellation of alternatives |
+| **Referral signals** | Unprompted sharing, introductions to colleagues, forwarded offers |
+| **Objection signals** | Price pushback, feature requests before committing, "let me think about it" patterns |
+| **Abandonment signals** | Drop-off points, cart abandonment, pricing page bounces, unsubscribes |
+
+### 3. Monitor for Contamination
+
+During execution, watch for these protocol violations:
+
+| Contamination Risk | Warning Sign | Corrective Action |
+|-------------------|--------------|-------------------|
+| **Selection bias drift** | Participants do not match screening criteria | Tighten recruitment; flag non-qualifying data |
+| **Novelty spike** | High early engagement that drops sharply | Extend test duration; weight later data more heavily |
+| **Observer effect** | Participants behave differently because they know they are being watched | Reduce visibility of observation; use passive measurement |
+| **Mid-test changes** | Temptation to tweak pricing, copy, or targeting during the test | Do not change anything. Log the urge but maintain protocol |
+| **Sample contamination** | Word-of-mouth spreading between test and non-test groups | Track referral sources; segment data if needed |
+
+### 4. Check Abort Criteria
+
+At each checkpoint, evaluate whether to continue or stop:
+
+| Checkpoint | Condition | Action |
+|------------|-----------|--------|
+| **25% of sample reached** | Zero conversions AND zero engagement signals | Consider aborting — fundamental value proposition may be wrong |
+| **50% of sample reached** | Conversion rate below failure threshold | Flag concern — continue to planned sample size if possible |
+| **75% of sample reached** | Conversion rate between failure and success thresholds | Continue — inconclusive results still need full sample |
+| **Full sample reached** | Any result | Stop recruiting. Move to analysis. |
+| **Any time** | External event invalidates test (competitor launch, market shift, PR crisis) | Abort, document reason, redesign if needed |
+
+### 5. Capture Qualitative Context
+
+While the quantitative data accumulates, capture qualitative signals that will enrich your analysis:
+
+| Category | What to Document |
+|----------|-----------------|
+| **Verbatim quotes** | What did participants say when they saw the price? When they decided to convert or not? |
+| **Objection patterns** | What reasons did non-converters give? Group them into categories. |
+| **Comparison behavior** | Did participants mention or compare to specific alternatives? |
+| **Unexpected behavior** | Anything you did not anticipate — positive or negative |
+| **Emotional reactions** | Enthusiasm, hesitation, confusion, frustration — these are data |
+
+---
+
+## Your Turn
+
+Launch the test, maintain the execution log, monitor for contamination, and capture both quantitative and qualitative signals. Share your progress as data accumulates, and I will help you interpret early signals without compromising the experiment.
+
+---
+
+**[a]** Advanced Elicitation — Deep dive into data interpretation, statistical significance, and mid-experiment decision-making
+**[p]** Party Mode — Bring in other Vortex agents to review your execution approach and early signals
+**[c]** Continue — Proceed to results analysis
+
+---
+
+## Next Step
+
+When your experiment has reached the planned sample size or an abort condition has been triggered, I will load:
+
+{project-root}/_bmad/bme/_vortex/workflows/proof-of-value/steps/step-05-analyze.md

package/_bmad/bme/_vortex/workflows/proof-of-value/steps/step-05-analyze.md

@@ -0,0 +1,116 @@
+---
+step: 5
+workflow: proof-of-value
+title: Analyze Results
+---
+
+# Step 5: Analyze Results
+
+The experiment is complete. Now let the data speak — and resist the urge to hear what you want to hear.
+
+## Why This Matters
+
+Analysis is where confirmation bias does the most damage. Teams that invested weeks designing and running a value test are psychologically invested in a positive result. The discipline here is radical honesty: compare actual results to pre-committed thresholds, surface inconvenient data, and let the market verdict stand. If you move the goalposts after seeing the data, you have learned nothing. The pre-committed success and failure thresholds from Steps 2 and 3 exist precisely for this moment — to prevent post-hoc rationalization from overriding market evidence.
+
+## Your Task
+
+### 1. Report Raw Results
+
+Before interpreting anything, lay out the raw data:
+
+**Quantitative Results:**
+
+| Metric | Pre-Committed Target | Pre-Committed Failure | Actual Result | Verdict |
+|--------|---------------------|-----------------------|---------------|---------|
+| **Primary metric** | *From Step 2* | *From Step 2* | *Actual observed value* | Met / Failed / Inconclusive |
+| **WTP metric** | *From Step 3* | *From Step 3* | *Actual observed value* | Met / Failed / Inconclusive |
+| **Secondary metric** | | | | |
+| **Guardrail metric** | | | | |
+
+**Execution Quality Check:**
+
+| Dimension | Assessment |
+|-----------|-----------|
+| **Sample size achieved** | Did you reach the planned sample size? If not, why? |
+| **Protocol adherence** | Were there any deviations from the designed protocol? |
+| **Contamination events** | Did any contamination risks materialize? How were they handled? |
+| **Duration** | Did the test run for the planned duration? |
+| **Data quality** | Is the data clean and reliable? Any gaps or anomalies? |
+
+### 2. Analyze Against Pre-Committed Thresholds
+
+Use the thresholds you set in Steps 2 and 3. Do not adjust them.
+
+**Value Hypothesis Verdict:**
+
+| Outcome | Criteria | Your Result |
+|---------|----------|-------------|
+| **Confirmed** | Primary metric meets or exceeds target threshold AND WTP metric meets target | The market is telling you there is a business here |
+| **Partially confirmed** | Some metrics met, others fell short OR results are in the inconclusive zone | Promising signal, but not strong enough to commit — refine and retest |
+| **Rejected** | Primary metric at or below failure threshold OR WTP below failure threshold | The market is telling you this value proposition does not work at this price for this segment |
+| **Inconclusive** | Results fall between success and failure thresholds | The test was not decisive — you need a better test, a larger sample, or a different method |
+
+### 3. Analyze Willingness-to-Pay Specifically
+
+Willingness-to-pay deserves its own analysis because it directly determines business viability:
+
+| WTP Dimension | Finding |
+|---------------|---------|
+| **Price sensitivity** | How did conversion change across price points tested? Where is the demand curve inflection? |
+| **Acceptable price range** | What price range showed the strongest conversion signal? |
+| **Value-to-price perception** | Did qualitative feedback suggest customers saw the price as fair, expensive, or cheap relative to value? |
+| **Competitive price position** | How did your conversion compare when priced above, at, or below competitor alternatives? |
+| **Segment variation** | Did willingness-to-pay vary significantly across sub-segments? |
+| **Payment friction** | Did customers who saw the price engage but drop off at the payment step? What does that signal? |
+
+### 4. Synthesize Qualitative Evidence
+
+Integrate the qualitative signals captured in Step 4:
+
+| Category | Key Findings |
+|----------|-------------|
+| **Strongest buy signals** | What did converters say or do that indicated genuine value perception? |
+| **Strongest rejection signals** | What did non-converters cite as the reason? Price? Value? Timing? Switching cost? |
+| **Objection patterns** | What were the top 3 objections? How many participants raised each? |
+| **Competitive mentions** | Which alternatives were mentioned? How were you compared? |
+| **Unexpected findings** | What did you learn that you did not expect? This is often the most valuable data. |
+| **Segment insights** | Did different segments behave differently? Who was most willing to pay? |
+
+### 5. Derive the Business Case Assessment
+
+Based on all evidence, assess the business case:
+
+| Dimension | Assessment |
+|-----------|-----------|
+| **Demand strength** | Is there sufficient demand at the target price to sustain a business? |
+| **Willingness-to-pay** | Will customers pay enough to cover costs and generate acceptable margin? |
+| **Unit economics** | Given conversion rates and pricing data, do the unit economics work? |
+| **Customer acquisition cost** | Based on test recruitment costs, what does CAC look like? Is it sustainable? |
+| **Competitive defensibility** | Given competitive alternatives, can you sustain the value proposition? |
+| **Scalability signal** | Do the results suggest this will scale beyond the test population? |
+
+### 6. Honest Assessment: What Did the Market Actually Tell You?
+
+Complete this sentence honestly:
+
+> **The market told us that** [target segment] [will / will not / might] pay [amount] for [value proposition] because [evidence-based rationale]. **The strongest evidence for this conclusion is** [specific data point]. **The biggest remaining uncertainty is** [what you still do not know].
+
+---
+
+## Your Turn
+
+Report the raw results, analyze them against pre-committed thresholds, synthesize the qualitative evidence, and produce the honest business case assessment. Share your analysis and I will help ensure you are reading the data truthfully — especially if the results are ambiguous.
+
+---
+
+**[a]** Advanced Elicitation — Deep dive into statistical analysis, segment breakdowns, and unit economics modeling
+**[p]** Party Mode — Bring in other Vortex agents to challenge your interpretation and look for blind spots
+**[c]** Continue — Proceed to documenting findings and routing
+
+---
+
+## Next Step
+
+When your analysis is complete and the business case assessment is honest, I will load:
+
+{project-root}/_bmad/bme/_vortex/workflows/proof-of-value/steps/step-06-document.md

package/_bmad/bme/_vortex/workflows/proof-of-value/steps/step-06-document.md

@@ -0,0 +1,147 @@
+---
+step: 6
+workflow: proof-of-value
+title: Document Findings & Route
+---
+
+# Step 6: Document Findings & Route
+
+Time to package everything into a durable artifact. The experiment is done, the analysis is complete, and the market has spoken. Now we produce the HC4 Experiment Context artifact so the evidence travels forward — and route to the next step in the Vortex.
+
+## Why This Matters
+
+Evidence that lives in your head or a chat thread is evidence that gets lost. The HC4 artifact captures what you tested, what you learned, and what the market told you in a format that any downstream agent or stakeholder can consume. Without this artifact, the next team member or Vortex agent starts from zero. With it, they start from validated evidence. Packaging the proof-of-value results as an HC4 also forces a final discipline check: can you document the findings without distorting them?
+
+## Your Task
+
+### 1. Review Your Findings
+
+Before packaging, do a final integrity check:
+
+| Check | Status |
+|-------|--------|
+| Raw results match the execution log from Step 4 | [ ] Verified |
+| Verdicts match pre-committed thresholds from Steps 2 and 3 — not post-hoc adjusted thresholds | [ ] Verified |
+| Willingness-to-pay analysis is based on behavioral signals, not stated preferences | [ ] Verified |
+| Qualitative evidence includes both confirming AND disconfirming signals | [ ] Verified |
+| Business case assessment is grounded in data, not optimism | [ ] Verified |
+
+### 2. Generate the HC4 Artifact
+
+I will produce the HC4 Experiment Context artifact with this structure:
+
+```yaml
+---
+contract: HC4
+type: artifact
+source_agent: wade
+source_workflow: proof-of-value
+target_agents: [noah]
+input_artifacts:
+  - path: "_bmad-output/vortex-artifacts/{your-upstream-artifact}"
+    contract: HC3
+created: YYYY-MM-DD
+---
+```
+
+**HC4 Required Body Sections:**
+
+1. **Experiment Summary** — Name, description, type (`Proof of Value`), duration, graduation status
+2. **Hypothesis Tested** — Value hypothesis statement, riskiest assumption, expected outcome, expected behavior change
+3. **Experiment Method** — Validation method used, sample size, duration, recruitment, controls
+4. **Pre-Defined Success Criteria** — Metric, target threshold, actual result, met/not met (from Steps 2-3)
+5. **Additional Results** — Supplementary quantitative metrics, qualitative findings, unexpected discoveries
+6. **Confirmed/Rejected Hypotheses** — Hypothesis status, assumption status, core learning sentence, conditions
+7. **Strategic Context** — Vortex stream, assumption tested, decision it informs, implications
+8. **Production Readiness** *(if graduating)* — Metrics to monitor, expected behavior, signal thresholds
+
+**Save to:** `{output_folder}/vortex-artifacts/hc4-experiment-context-{date}.md`
+
+I will create this file with all sections populated from your work in Steps 1-5 once you confirm the content is ready.
+
+### 3. Determine the Graduation Status
+
+Based on your analysis, select the appropriate status:
+
+| Status | Criteria | Meaning |
+|--------|----------|---------|
+| **Graduated** | Value hypothesis confirmed, WTP validated, unit economics work | This value proposition is ready for production investment (MVP or scaling) |
+| **Completed** | Experiment ran fully, results captured, but not graduating | Valuable learning captured; value proposition needs refinement, pivot, or further testing |
+| **Terminated** | Experiment stopped early due to abort criteria | The market signal was clear enough to stop — typically a strong rejection |
+
+### 4. Write the Core Learning Statement
+
+Summarize the entire proof-of-value in one sentence:
+
+> "We [validated / invalidated / partially validated] that [target segment] will [pay $X / change behavior Y] for [value proposition], discovering that [key insight about market demand, pricing, or competitive position]."
+
+This sentence is the most important output of the entire workflow. It should be precise enough that someone who reads only this sentence understands the market verdict.
+
+### 5. Final Validation Questions
+
+Before we finalize the HC4 artifact:
+
+**Evidence Integrity:**
+- [ ] Would you be comfortable presenting these findings to a skeptical investor?
+- [ ] If someone re-ran this exact experiment, would they likely reach the same conclusion?
+- [ ] Have you documented disconfirming evidence as prominently as confirming evidence?
+
+**Completeness:**
+- [ ] Are all pre-defined success criteria evaluated with actual results?
+- [ ] Is the willingness-to-pay analysis specific enough to inform pricing decisions?
+- [ ] Are qualitative findings included with verbatim quotes where available?
+
+**Routing Readiness:**
+- [ ] Is the graduation status justified by the data, not by organizational pressure?
+- [ ] Is the core learning statement honest and falsifiable?
+- [ ] Does the strategic context section make clear what decision this evidence enables?
+
+---
+
+## Your Turn
+
+Review your findings, confirm readiness for the HC4 artifact, determine graduation status, and write the core learning statement. When you are ready, I will generate the final HC4 artifact.
+
+---
+
+## Vortex Compass
+
+Based on what you just completed, here are your evidence-driven options:
+
+| If you learned... | Consider next... | Agent | Why |
+|---|---|---|---|
+| Value hypothesis confirmed, WTP validated, business case is strong | mvp | Wade | Business value is proven — design and build the minimum viable product (HC4 feeds MVP workflow) |
+| Value confirmed but need production signal monitoring set up | production-monitoring | Noah | Set up monitoring for production metrics identified in HC4 Section 8 |
+| Value confirmed — capture the learning for stakeholder communication | learning-card | Max | Package the proof-of-value evidence as a learning card for decision-makers |
+| Value partially confirmed — pricing needs refinement | proof-of-value | Wade | Re-run Steps 3-5 with adjusted pricing hypothesis |
+| Value unclear — need deeper understanding of customer segment | user-interview | Isla | Uncover why willingness-to-pay was weaker than expected (HC9 routing) |
+| Value rejected — need to understand whether to pivot, patch, or kill | pivot-patch-persevere | Max | Make a structured decision about the value proposition's future |
+| Value rejected — problem definition may be wrong | research-convergence | Mila | Re-examine the underlying problem; the pain may not be what you thought |
+| Technical assumptions changed during value testing | proof-of-concept | Wade | New technical risks surfaced that need feasibility validation |
+
+> **Note:** These are evidence-based recommendations. You can navigate to any Vortex agent
+> at any time based on your judgment.
+
+**Or run Max's [VN] Vortex Navigation** for a full gap analysis across all streams.
+
+### Insufficient Evidence for Routing
+
+If the evidence gathered so far does not clearly point to a single next step:
+
+| To route to... | You need... |
+|----------------|-------------|
+| Wade (MVP) | Confirmed value hypothesis, validated WTP, positive unit economics signal |
+| Wade (re-run PoV) | Partially confirmed value with specific pricing or segment hypothesis to retest |
+| Noah | Graduated experiment with production readiness section completed in HC4 |
+| Isla | Specific question about customer behavior or segment that quantitative data could not answer |
+| Max (learning-card) | Completed HC4 artifact with clear core learning statement |
+| Max (pivot-patch-persevere) | Rejected or partially confirmed hypothesis with enough evidence to inform the pivot decision |
+| Mila | Evidence that the underlying problem definition is flawed, not just the value proposition |
+
+**Workflow-specific signals:**
+- WTP data is ambiguous across all price points tested → consider re-running **Step 3** with a different pricing model or method
+- Conversion varied dramatically across segments → consider routing to **Isla** for deeper segment research before committing to a segment
+- Qualitative objections point to a different value proposition entirely → consider routing back to **Step 1** to redefine the value hypothesis
+- Results were inconclusive due to insufficient sample → consider re-running **Step 4** with stronger recruitment
+
+**Recommended:** If the path forward is unclear, run **Max's [VN] Vortex Navigation** for a full gap analysis across all streams.

package/_bmad/bme/_vortex/workflows/proof-of-value/validate.md

@@ -0,0 +1,30 @@
+# Validate Proof of Value
+
+**Status:** Coming in v1.2.0
+
+**Agent:** Wade (Lean Experiments Specialist)
+
+**Stream:** Externalize
+
+## Overview
+
+This validation workflow helps you review proof-of-values to ensure they test real user value, not just feature interest.
+
+## What Gets Validated
+
+- Is the value hypothesis specific and testable?
+- Are target users clearly defined?
+- Is the validation approach appropriate?
+- Are results based on actual user feedback?
+- Is the business case justified by evidence?
+- Does this test value perception, not just feature desirability?
+
+## Coming in v1.2.0
+
+This validation workflow will be available alongside the proof-of-value workflow in March 2026.
+
+## Questions?
+
+For questions or to request early access:
+- GitHub Issues: https://github.com/amalik/convoke-agents/issues
+- Tag with: `workflow:proof-of-value` and `v1.2.0`

package/_bmad/bme/_vortex/workflows/proof-of-value/workflow.md

@@ -0,0 +1,26 @@
+---
+workflow: proof-of-value
+type: step-file
+description: Validate business value and market demand
+author: Wade (lean-experiments-specialist)
+version: 1.2.0
+---
+
+# Proof-of-Value Workflow
+
+Validate that you SHOULD build this (business case) after proving you CAN.
+
+## Steps Overview
+
+1. **Define Value Hypothesis** - What value does this create?
+2. **Design Value Test** - How will you validate demand?
+3. **Run Market Experiment** - Test with real users/customers
+4. **Measure Willingness to Pay** - Will they pay? How much?
+5. **Calculate Business Case** - Is this viable?
+6. **Document Decision** - Build, pivot, or kill?
+
+---
+
+## INITIALIZATION
+
+Load step: {project-root}/_bmad/bme/_vortex/workflows/proof-of-value/steps/step-01-value-hypothesis.md

package/_bmad/bme/_vortex/workflows/research-convergence/steps/step-01-setup.md

@@ -0,0 +1,69 @@
+---
+step: 1
+workflow: research-convergence
+title: Setup & Input Validation
+---
+
+# Step 1: Setup & Input Validation
+
+Before we synthesize anything, we need to know exactly what research we're working with and verify it's ready for convergence.
+
+## Why This Matters
+
+Three patterns converge on this insight: the quality of your problem definition is directly tied to the quality and completeness of the research feeding it. Garbage in, garbage out. This step ensures we start with solid foundations by validating your input artifacts against the HC1 schema — or guiding you to provide what's needed if your research comes from outside the Vortex pattern.
+
+## Your Task
+
+### 1. What Artifacts Do You Have?
+
+Mila expects research artifacts — ideally produced by Isla's workflows as HC1-compliant empathy artifacts:
+- **Empathy maps** (from Isla's `empathy-map` workflow)
+- **Interview syntheses** (from Isla's `user-interview` workflow)
+- **Observation reports** (from Isla's `user-discovery` workflow)
+
+You can also bring **any well-formed research input** — Mila accepts artifacts from outside the Vortex pattern. They don't have to be HC1-compliant, but having structured research makes convergence stronger.
+
+### 2. List Your Input Artifacts
+
+Please provide the file paths or describe the artifacts you want to synthesize. For example:
+- `_bmad-output/vortex-artifacts/hc1-empathy-map-busy-parents-2026-02-20.md`
+- `_bmad-output/vortex-artifacts/hc1-interview-synthesis-2026-02-21.md`
+- Or: "I have interview notes in a Google Doc and survey results in a spreadsheet"
+
+### 3. Input Validation
+
+I'll check each artifact against the HC1 schema to assess readiness:
+
+**HC1 Frontmatter Check:**
+- `contract: HC1`
+- `type: artifact`
+- `source_agent` (who produced it)
+- `source_workflow` (which workflow)
+- `target_agents: [mila]`
+- `input_artifacts` (upstream references)
+- `created` (date)
+
+**HC1 Body Section Check:**
+- Executive Summary
+- Research Context
+- Synthesized Insights
+- Key Themes
+- Pain Points
+- Desired Gains
+- Recommendations
+
+**If artifacts are non-conforming:** That's okay — we don't reject research. I'll guide you to identify which sections are present and which gaps we need to work around during synthesis. The data shows that even partial research can produce useful convergence when we're explicit about what evidence we have and what we're missing.
+
+> For the full HC1 schema reference, see `{project-root}/_bmad/bme/_vortex/contracts/hc1-empathy-artifacts.md`
+
+---
+
+## Your Turn
+
+Please list your input artifacts — file paths, descriptions, or both. I'll validate them and we'll proceed to analysis.
+
+## Next Step
+
+When your artifacts are listed and validated, I'll load:
+
+{project-root}/_bmad/bme/_vortex/workflows/research-convergence/steps/step-02-context.md

package/_bmad/bme/_vortex/workflows/research-convergence/steps/step-02-context.md

@@ -0,0 +1,70 @@
+---
+step: 2
+workflow: research-convergence
+title: Context Loading & Analysis
+---
+
+# Step 2: Context Loading & Analysis
+
+Now that we know what artifacts we're working with, let's load them all and start identifying what the research is collectively telling us.
+
+## Why This Matters
+
+Individual research artifacts each tell part of the story. The data shows that the real insights emerge when you lay them side by side and look for where they agree, where they disagree, and where they're silent. Cross-artifact analysis is where convergence begins.
+
+## Your Task
+
+### 1. Load All Input Artifacts
+
+Read through each of your input artifacts completely. Don't skim — the details matter for synthesis.
+
+For each artifact, note:
+- **Who** was researched (user segments, personas, interview subjects)
+- **What** was discovered (key findings, insights, themes)
+- **How strong** the evidence is (direct quotes, observed behavior, survey data, or assumptions)
+
+### 2. Cross-Artifact Analysis
+
+Now look across all your artifacts together. Here's what the research is telling us — or rather, here's how to listen:
+
+**Themes:** What topics or concerns appear in multiple artifacts?
+- Example: "Both the empathy map and interview synthesis mention time pressure as a recurring theme"
+
+**Patterns:** What behaviors or needs show up repeatedly?
+- Example: "Three separate artifacts reference users creating workarounds for the same limitation"
+
+**Contradictions:** Where do artifacts disagree or present conflicting evidence?
+- Example: "The empathy map suggests users want more features, but interviews reveal they feel overwhelmed by current complexity"
+
+**Evidence Gaps:** What questions remain unanswered across all artifacts?
+- Example: "We have strong evidence about pain points but almost no data on what users would consider a successful outcome"
+
+### 3. Synthesis Inventory
+
+Based on your cross-artifact analysis, build a mental model of what the artifacts collectively tell us. You don't need to write a formal document here — the HC2 problem definition comes in Step 5. Instead, organize your thinking:
+
+- **Strongest signal:** What finding has the most evidence behind it?
+- **Most surprising insight:** What did you not expect to find?
+- **Biggest gap:** What do you wish you knew but don't?
+- **Emerging problem shape:** If you had to describe the core problem in one sentence right now, what would it be?
+
+---
+
+## Example Cross-Artifact Analysis
+
+**From 3 artifacts** (empathy map, interview synthesis, observation report):
+- **Theme:** Time pressure appears in all 3 artifacts — users consistently mention feeling rushed
+- **Pattern:** 2 of 3 artifacts show users building personal workarounds (spreadsheets, sticky notes) for the same data consolidation problem
+- **Contradiction:** Empathy map suggests users want more features; interviews reveal they feel overwhelmed by current complexity
+- **Gap:** Strong evidence on pain points, but almost no data on what "success" looks like to users
+- **Emerging problem shape:** "Users spend more time gathering data than acting on it, and they've given up asking for help"
+
+## Your Turn
+
+Walk me through your cross-artifact analysis. What themes, patterns, contradictions, and gaps do you see? Three patterns converge on this insight — share what you're finding and I'll help you sharpen it.
+
+## Next Step
+
+When we've mapped the cross-artifact landscape, I'll load:
+
+{project-root}/_bmad/bme/_vortex/workflows/research-convergence/steps/step-03-jtbd-framing.md

package/_bmad/bme/_vortex/workflows/research-convergence/steps/step-03-jtbd-framing.md

@@ -0,0 +1,81 @@
+---
+step: 3
+workflow: research-convergence
+title: Jobs-to-be-Done Framing
+---
+
+# Step 3: Jobs-to-be-Done Framing
+
+Time to take everything we've analyzed and frame it through the Jobs-to-be-Done lens. This is where scattered findings become an actionable problem statement.
+
+## Why This Matters
+
+JTBD framing turns observations into actionable problem statements. Instead of listing what users do or feel, we define what they're trying to accomplish — and that's what drives solutions. Across all our research, the core problem is always rooted in an unmet job.
+
+## Your Task
+
+### 1. Frame the Primary JTBD
+
+Using the evidence from your artifacts, complete this structure:
+
+> **When** [situation — the triggering context],
+> **I want to** [motivation — what the user wants to accomplish],
+> **so I can** [expected outcome — the desired end state].
+
+**Tips for Strong JTBD Framing:**
+- The **situation** should come directly from your research (observed contexts, interview triggers)
+- The **motivation** should reflect what users actually said or did, not what we assume they want
+- The **expected outcome** should describe a changed state, not a feature
+
+**Example:**
+> **When** I'm managing multiple marketing campaigns across 6 different tools,
+> **I want to** see all campaign performance in one place,
+> **so I can** make real-time optimization decisions without wasting 2 hours on data consolidation.
+
+### 2. Extract Job Types
+
+From the artifact evidence, identify:
+
+**Functional Job** *(required)*
+The practical task the user is trying to accomplish.
+- What observable action are they performing?
+- What's the measurable outcome they need?
+
+**Emotional Job** *(if evidence supports)*
+How the user wants to feel while doing the job.
+- What emotions surfaced in interviews or observations?
+- What anxieties or frustrations did the research capture?
+
+**Social Job** *(if evidence supports)*
+How the user wants to be perceived by others.
+- What social dynamics emerged from the research?
+- How does the user's role or reputation connect to this job?
+
+### 3. Ground Every Claim in Evidence
+
+For each element of your JTBD, note which artifact supports it:
+- "Situation comes from empathy map (artifact 1) — user context section"
+- "Motivation from interview synthesis (artifact 2) — direct quote: '...'"
+- "Outcome from observation report (artifact 3) — observed behavior"
+
+The data shows that JTBD statements grounded in specific evidence are far more useful than those based on intuition.
+
+---
+
+## Your Turn
+
+Draft your Primary JTBD statement and identify the Functional, Emotional, and Social jobs. Link each to specific artifact evidence.
+
+---
+
+**[a]** Advanced Elicitation — Deep dive into JTBD framing with guided questioning
+**[p]** Party Mode — Bring in other Vortex agents for collaborative discussion
+**[c]** Continue — Move to Pains & Gains analysis
+
+---
+
+## Next Step
+
+When your JTBD is framed and grounded in evidence, I'll load:
+
+{project-root}/_bmad/bme/_vortex/workflows/research-convergence/steps/step-04-pains-gains.md