dojo.md 0.1.0

package/courses/competitive-battlecards/scenarios/level-3/category-creation-battlecards.yaml

@@ -0,0 +1,66 @@
+meta:
+  id: category-creation-battlecards
+  level: 3
+  course: competitive-battlecards
+  type: output
+  description: "Create battlecards for category creation — position your product as a new category rather than competing in an existing one where incumbents dominate"
+  tags: [sales, battlecards, category-creation, positioning, strategy, advanced]
+
+state: {}
+
+trigger: |
+  You're the VP of Product Marketing at "Synapse" (AI-powered
+  revenue intelligence platform, $45M ARR). You're stuck in the
+  "conversation intelligence" category alongside Gong ($250M ARR)
+  and Chorus ($100M ARR, now part of ZoomInfo). You can't win a
+  feature comparison against Gong — they have 5x your R&D budget.
+
+  Your CEO's strategy: "Stop competing in conversation intelligence.
+  Create the 'Revenue Intelligence' category where we're the
+  leader." Synapse does more than call recording — you analyze
+  pipeline data, CRM activity, email sentiment, and market signals
+  to predict deal outcomes. Gong and Chorus focus on calls.
+
+  The challenge: Your reps keep getting pulled into conversation
+  intelligence comparisons. Prospects say "we're evaluating Gong
+  and Synapse" — framing you as a Gong alternative. Your battlecards
+  need to REFRAME the evaluation, not just compete within it.
+
+  Tasks:
+  1. Create the "Category Creation Battlecard" — a fundamentally
+     different type of battlecard that doesn't compare features
+     but redefines the evaluation. Include: the category narrative
+     (why "conversation intelligence" is limited), the new category
+     definition ("revenue intelligence" — what it includes that CI
+     doesn't), the positioning against Gong that's NOT a comparison
+     but a reframe, and the questions reps ask to shift the
+     prospect's frame of reference.
+
+  2. Write the "Evaluation Reframe Talk Track" — the specific words
+     reps use when a prospect says "we're evaluating you against
+     Gong." The talk track must: not disparage Gong, acknowledge
+     their strength in conversation intelligence, and expand the
+     evaluation criteria to include pipeline intelligence, deal
+     prediction, and revenue forecasting where Gong doesn't compete.
+
+  3. Address the risk: category creation can confuse prospects who
+     came looking for a specific solution. Write the "When to
+     Compete vs. When to Create" decision framework — situations
+     where you should compete head-to-head on conversation
+     intelligence (the prospect only wants call recording) vs.
+     when you should reframe to revenue intelligence (the prospect
+     wants broader revenue visibility).
+
+assertions:
+  - type: llm_judge
+    criteria: "Category creation battlecard reframes rather than compares — defines the new category compellingly, positions Gong as a subset rather than a competitor, and provides questions that naturally expand evaluation criteria"
+    weight: 0.35
+    description: "Reframing category battlecard"
+  - type: llm_judge
+    criteria: "Evaluation reframe talk track is natural — acknowledges Gong's strength, expands the conversation without disparaging, and provides a credible reason to evaluate on broader criteria"
+    weight: 0.35
+    description: "Natural evaluation reframe"
+  - type: llm_judge
+    criteria: "Compete vs. create decision framework is practical — identifies clear signals for when to compete (narrow need) vs. reframe (broad need), preventing category creation confusion in prospects who just want a simple solution"
+    weight: 0.30
+    description: "Practical compete vs create framework"

package/courses/competitive-battlecards/scenarios/level-3/competitive-deal-strategy.yaml

@@ -0,0 +1,66 @@
+meta:
+  id: competitive-deal-strategy
+  level: 3
+  course: competitive-battlecards
+  type: output
+  description: "Use battlecards for competitive deal strategy — move beyond reactive objection handling to proactive competitive deal planning"
+  tags: [sales, battlecards, deal-strategy, proactive, planning, advanced]
+
+state: {}
+
+trigger: |
+  You're a Strategic Account Manager at "InnoSec" (cybersecurity
+  platform, $90M ARR). You've just learned that your biggest prospect
+  (Fortune 500 bank, potential $2M deal) is running a formal
+  evaluation with 4 vendors: you, CrowdStrike, Palo Alto, and
+  Microsoft Defender.
+
+  The evaluation timeline: 16 weeks
+  Your champion: CISO, recently hired, wants to make a bold choice
+  Your competitor intelligence:
+  - CrowdStrike: Already has 2 departments in this bank. Incumbent
+    advantage. Their enterprise AE has been selling here for 3 years.
+  - Palo Alto: The bank's CTO is a former Palo Alto VP. Strong
+    relationship at the top.
+  - Microsoft Defender: "Free" with existing Microsoft E5 license.
+    The CFO is pushing this as the "no additional cost" option.
+
+  Your battlecards tell you WHAT each competitor's strengths and
+  weaknesses are. But for a $2M deal, you need a competitive DEAL
+  STRATEGY — how to use competitive intelligence proactively across
+  16 weeks.
+
+  Tasks:
+  1. Create the "Competitive Deal Strategy Map" for this specific
+     deal. For each of the 16 weeks, identify: the primary
+     competitive threat, your competitive move, the battlecard
+     content you deploy, and the risk if you don't act. This is
+     a chess game — you need to think 3 moves ahead.
+
+  2. Write the "Competitor Neutralization Plan" for each competitor:
+     - CrowdStrike: How do you neutralize their incumbent advantage?
+     - Palo Alto: How do you neutralize the CTO's personal
+       relationship?
+     - Microsoft Defender: How do you neutralize the "free" argument?
+     Each plan must be specific, ethical, and actionable within the
+     evaluation timeline.
+
+  3. Design the "Competitive Win Room" — a cross-functional war room
+     for this specific deal. Who needs to be involved (AE, SE, exec
+     sponsor, CI analyst, product), what each person's competitive
+     role is, the meeting cadence, and the intelligence-gathering
+     assignments for each team member.
+
+assertions:
+  - type: llm_judge
+    criteria: "Deal strategy map is proactive — 16-week plan shows specific competitive moves timed to evaluation phases, anticipates competitor actions, and positions proactively rather than reacting to competitor tactics"
+    weight: 0.35
+    description: "Proactive deal strategy map"
+  - type: llm_judge
+    criteria: "Neutralization plans are specific and ethical — addresses CrowdStrike's incumbency (redefine requirements), Palo Alto's relationship (earn CISO's trust through value), Microsoft's 'free' (expose hidden costs) with actionable, ethical tactics"
+    weight: 0.35
+    description: "Specific ethical neutralization plans"
+  - type: llm_judge
+    criteria: "Win room design is operationally realistic — cross-functional team with clear competitive roles, realistic meeting cadence, and specific intelligence-gathering assignments that add value without being busywork"
+    weight: 0.30
+    description: "Realistic win room design"

package/courses/competitive-battlecards/scenarios/level-3/competitive-intelligence-program.yaml

@@ -0,0 +1,75 @@
+meta:
+  id: competitive-intelligence-program
+  level: 3
+  course: competitive-battlecards
+  type: output
+  description: "Build a comprehensive competitive intelligence program — create the organizational capability that feeds battlecards with continuous, validated intelligence"
+  tags: [sales, battlecards, competitive-intelligence, program, organization, advanced]
+
+state: {}
+
+trigger: |
+  You're the newly hired Director of Competitive Intelligence at
+  "TechBridge" (integration platform, $95M ARR, 80 sales reps).
+  You're the company's first CI hire. Currently, competitive
+  intelligence is scattered: marketing has some competitor info,
+  product has feature comparisons, and individual reps have tribal
+  knowledge that they don't share.
+
+  Your mandate: Build TechBridge's competitive intelligence program
+  from scratch. Your budget: $200K/year (1 FTE — you, plus tools
+  and resources).
+
+  Current state:
+  - No CI tools — monitoring is manual
+  - Battlecards: 3 exist (outdated, created by a former PMM)
+  - Win/loss analysis: None formal, occasional anecdotes
+  - Competitive data sources: Reps' personal knowledge, competitor
+    websites, occasional Gartner reports
+  - Cross-functional alignment: Zero — product, sales, and marketing
+    have different views of the competitive landscape
+  - Competitors tracked: 8 (but should probably be 4-5 priority)
+  - Sales feedback: No mechanism for reps to share competitive intel
+
+  Key competitors by revenue impact:
+  1. MuleSoft (Salesforce): 35% of lost competitive deals
+  2. Boomi (Dell): 25% of lost competitive deals
+  3. Workato: 20% of lost competitive deals
+  4. Tray.io: 10% of lost competitive deals
+  5-8. Long tail (10% combined)
+
+  Tasks:
+  1. Design the "CI Program Launch Plan" — your first 90 days. Break
+     it into 3 phases:
+     Phase 1 (Days 1-30): Quick wins — what you deliver immediately
+     Phase 2 (Days 31-60): Systems — what infrastructure you build
+     Phase 3 (Days 61-90): Scale — what ongoing processes you establish
+     Each phase must have deliverables, cross-functional dependencies,
+     and success metrics.
+
+  2. Create the "Competitive Intelligence Technology Stack" — with
+     $200K budget (minus your salary), recommend the tools you'd
+     invest in. Evaluate: CI platforms (Crayon, Klue, Kompyte),
+     monitoring tools, CRM integration, and content distribution.
+     Justify each investment with expected ROI.
+
+  3. Write the "Cross-Functional CI Charter" — the document that
+     defines how product, marketing, sales, and customer success
+     each contribute to and benefit from the CI program. Include:
+     each team's CI responsibilities, what they receive in return,
+     the meeting cadence, and the escalation path for competitive
+     emergencies.
+
+assertions:
+  - type: llm_judge
+    criteria: "90-day plan delivers quick wins first — Phase 1 produces immediately usable deliverables (top 4 battlecards, basic monitoring), Phase 2 builds infrastructure (tools, processes, feedback loops), Phase 3 establishes sustainable operations"
+    weight: 0.35
+    description: "Quick-win-first 90-day plan"
+  - type: llm_judge
+    criteria: "Technology stack is budget-appropriate — recommends tools that fit within the $200K budget (after salary), with clear ROI justification for each investment, not just listing every CI tool available"
+    weight: 0.30
+    description: "Budget-appropriate tech stack"
+  - type: llm_judge
+    criteria: "Cross-functional charter is realistic — defines specific responsibilities for each team (not just 'share information'), what each team gets in return, and a meeting cadence that busy teams will actually follow"
+    weight: 0.35
+    description: "Realistic cross-functional charter"

package/courses/competitive-battlecards/scenarios/level-3/competitive-pricing-strategy.yaml

@@ -0,0 +1,72 @@
+meta:
+  id: competitive-pricing-strategy
+  level: 3
+  course: competitive-battlecards
+  type: output
+  description: "Build pricing-focused battlecard content — create the TCO analyses, ROI models, and pricing positioning that help reps win price-sensitive competitive deals"
+  tags: [sales, battlecards, pricing, TCO, ROI, value, advanced]
+
+state: {}
+
+trigger: |
+  You're a Pricing Strategy Manager at "CloudGuard" (cloud security,
+  $70M ARR). Your pricing is $85/endpoint/month. Your competitors:
+  - CrowdStrike: $110/endpoint/month (enterprise premium)
+  - SentinelOne: $70/endpoint/month (your closest price competitor)
+  - Carbon Black (VMware): $45/endpoint/month (budget option)
+  - Open-source alternatives: Free (ClamAV, OSSEC)
+
+  Your pricing position: mid-range. You're cheaper than CrowdStrike
+  but more expensive than SentinelOne and Carbon Black. Prospects
+  constantly ask "why should I pay $85 when I can get SentinelOne
+  for $70 or Carbon Black for $45?"
+
+  Pricing data you have:
+  - Average deployment cost: CloudGuard $15K, CrowdStrike $45K,
+    SentinelOne $20K, Carbon Black $30K (requires more customization)
+  - Annual maintenance cost: CloudGuard $0 (included), CrowdStrike
+    $12K, SentinelOne $5K, Carbon Black $15K
+  - Average time to deploy: CloudGuard 2 weeks, CrowdStrike 8 weeks,
+    SentinelOne 3 weeks, Carbon Black 6 weeks
+  - False positive rate: CloudGuard 2%, CrowdStrike 1%, SentinelOne
+    5%, Carbon Black 8% (each 1% = $8K/year in analyst time)
+  - 3-year TCO for 500 endpoints (all-in):
+    CloudGuard: $1.53M
+    CrowdStrike: $2.38M
+    SentinelOne: $1.47M
+    Carbon Black: $1.71M
+
+  Tasks:
+  1. Create the "Pricing Battlecard Section" for each competitor.
+     Each must include: the per-unit price comparison (honest), the
+     TCO comparison (where you win against CrowdStrike and Carbon
+     Black), the value comparison (when per-unit price is the wrong
+     metric), and the specific pricing objection handler.
+
+  2. Build the "TCO Calculator Talk Track" — the 3-minute
+     conversation where reps walk prospects through total cost of
+     ownership. Include: the opening question, the data points to
+     present, the visual framework (how to sketch this on a
+     whiteboard or share screen), and the closing that connects
+     TCO to the prospect's budget.
+
+  3. Write the "Pricing Positioning Decision Matrix" — when to
+     compete on price (vs. Carbon Black: show TCO), when to compete
+     on value (vs. SentinelOne: show better outcomes), and when to
+     compete on premium (vs. CrowdStrike: show same quality, lower
+     price). Include the diagnostic questions that tell you which
+     pricing conversation to have.
+
+assertions:
+  - type: llm_judge
+    criteria: "Pricing battlecard sections are competitor-specific — CrowdStrike section shows TCO advantage, SentinelOne section pivots to value, Carbon Black section exposes hidden costs, each with honest per-unit comparison and specific objection handler"
+    weight: 0.35
+    description: "Competitor-specific pricing sections"
+  - type: llm_judge
+    criteria: "TCO calculator talk track is executable — 3-minute conversation with specific opening, data points, visual framework, and closing that a rep could actually deliver during a call"
+    weight: 0.30
+    description: "Executable TCO talk track"
+  - type: llm_judge
+    criteria: "Pricing decision matrix is clear — provides distinct strategies for each competitive situation (TCO, value, premium) with diagnostic questions that tell reps which conversation to have"
+    weight: 0.35
+    description: "Clear pricing decision matrix"

package/courses/competitive-battlecards/scenarios/level-3/competitive-review-sites.yaml

@@ -0,0 +1,70 @@
+meta:
+  id: competitive-review-sites
+  level: 3
+  course: competitive-battlecards
+  type: output
+  description: "Leverage review site intelligence for battlecards — mine G2, Capterra, and TrustRadius for competitive insights and integrate them into your battlecard strategy"
+  tags: [sales, battlecards, reviews, G2, Capterra, social-proof, advanced]
+
+state: {}
+
+trigger: |
+  You're a CI Manager at "CloudHR" (HR platform, $55M ARR). Review
+  sites are a goldmine of competitive intelligence, but your team
+  hasn't systematically used them. You have access to G2 (4,200
+  reviews across 5 competitors), Capterra (2,800 reviews), and
+  TrustRadius (950 reviews).
+
+  Your competitive landscape on G2:
+  - CloudHR: 4.3 stars (380 reviews), ranked #5 in category
+  - Workday: 4.0 stars (1,200 reviews), ranked #1 (volume matters)
+  - BambooHR: 4.5 stars (2,100 reviews), ranked #2
+  - Rippling: 4.8 stars (580 reviews), ranked #3
+  - Gusto: 4.3 stars (1,200 reviews), ranked #4
+
+  Key review insights you've extracted:
+  - CloudHR's top positive theme: "Best customer support" (mentioned
+    in 68% of positive reviews)
+  - CloudHR's top negative theme: "Limited reporting" (mentioned in
+    42% of negative reviews)
+  - Rippling's hidden weakness: "Terrible for companies over 500
+    employees" (mentioned in 35% of negative reviews, not visible
+    from their marketing)
+  - Workday's cost complaint: "You need consultants for everything"
+    (mentioned in 55% of negative reviews)
+
+  Tasks:
+  1. Create the "Review Site Intelligence Protocol" — how you
+     systematically extract competitive insights from review sites.
+     Include: what to look for (themes, not individual reviews),
+     how to quantify findings (thematic analysis), how often to
+     refresh (quarterly minimum), and how to validate review data
+     (incentivized reviews, competitor manipulation detection).
+
+  2. Write the battlecard sections powered by review site data.
+     For each competitor, create: the "What Their Customers Say"
+     section (citing real review themes), the "Hidden Weaknesses"
+     section (things you learn from reviews that aren't in their
+     marketing), and the "Social Proof Comparison" (how to use your
+     review positioning in sales conversations).
+
+  3. Write the "G2 Strategy" — how to improve your own G2 ranking
+     to support competitive selling. Include: ethical review
+     generation (asking happy customers, not incentivizing), response
+     strategy for negative reviews, and how to use G2 badges and
+     reports in battlecards. Address the ethics: what's acceptable
+     review management vs. review manipulation.
+
+assertions:
+  - type: llm_judge
+    criteria: "Intelligence protocol is systematic — provides a replicable method for extracting thematic insights from review sites, with quantification, refresh cadence, and validation for fake or incentivized reviews"
+    weight: 0.35
+    description: "Systematic review intelligence protocol"
+  - type: llm_judge
+    criteria: "Review-powered battlecard sections add unique value — competitor sections cite real review themes that aren't in marketing materials, providing intelligence reps can't get from competitor websites"
+    weight: 0.35
+    description: "Unique review-powered sections"
+  - type: llm_judge
+    criteria: "G2 strategy is ethical — review generation approach respects platform rules, addresses negative reviews constructively, and draws clear line between legitimate review management and manipulation"
+    weight: 0.30
+    description: "Ethical G2 strategy"

package/courses/competitive-battlecards/scenarios/level-3/cross-functional-battlecards.yaml

@@ -0,0 +1,75 @@
+meta:
+  id: cross-functional-battlecards
+  level: 3
+  course: competitive-battlecards
+  type: output
+  description: "Build battlecards with cross-functional input — integrate insights from product, customer success, marketing, and engineering into competitive content"
+  tags: [sales, battlecards, cross-functional, collaboration, product, CS, advanced]
+
+state: {}
+
+trigger: |
+  You're the Senior CI Analyst at "DataFlow" (data pipeline SaaS,
+  $60M ARR). Your battlecards have been good but one-dimensional —
+  built entirely from sales perspective. Your CRO says: "Product
+  knows things sales doesn't. CS hears things product doesn't. We
+  need cross-functional battlecards."
+
+  Cross-functional intelligence you need to integrate:
+
+  From Product team:
+  - Detailed technical comparison (API performance, architecture
+    differences, scalability benchmarks) that reps can't assess
+  - Roadmap items that will close competitive gaps in 3-6 months
+  - Technical limitations your competitors have that aren't visible
+    in demos but matter in production
+
+  From Customer Success:
+  - Why customers considered switching to competitors (churn risk
+    intelligence — "they evaluated Airbyte because of...")
+  - What existing customers love that you should emphasize more
+  - Real implementation timelines vs. competitor promises
+
+  From Marketing:
+  - How competitors position themselves (messaging analysis)
+  - What keywords competitors are targeting (SEO competitive data)
+  - What content resonates with prospects evaluating your category
+
+  From Engineering:
+  - Security and compliance differences (real, not marketing claims)
+  - Performance benchmarks (how your system actually performs under
+    load vs. competitors)
+  - Integration depth (surface-level vs. deep API integration)
+
+  Tasks:
+  1. Design the "Cross-Functional Battlecard Intelligence System" —
+     how each team contributes to battlecards. For each team: what
+     intelligence they provide (specific, not vague), the collection
+     mechanism (meeting, Slack channel, quarterly review), and how
+     their input gets translated into rep-usable battlecard content.
+
+  2. Write the battlecard section that comes from each team. Create
+     one section sourced from Product (technical differentiation),
+     one from CS (customer-validated proof points), one from
+     Marketing (positioning and messaging), and one from Engineering
+     (security/performance claims). Show how each section is
+     different from what sales alone would produce.
+
+  3. Address the coordination challenge: "Getting 4 teams to
+     contribute regularly is hard. They all think it's someone else's
+     job." Write the incentive structure and accountability framework
+     that makes cross-functional CI contribution sustainable.
+
+assertions:
+  - type: llm_judge
+    criteria: "Intelligence system has specific contributions per team — each team provides distinct, named intelligence types with clear collection mechanisms, not vague 'share competitive info' mandates"
+    weight: 0.35
+    description: "Specific team contributions"
+  - type: llm_judge
+    criteria: "Team-sourced sections are distinctly valuable — Product section has technical depth sales can't provide, CS section has real customer validation, Marketing section has messaging intelligence, Engineering section has verified performance claims"
+    weight: 0.30
+    description: "Distinctly valuable team sections"
+  - type: llm_judge
+    criteria: "Coordination framework is sustainable — incentive structure and accountability make cross-functional contribution feel like part of each team's job (not extra work), with realistic expectations for time commitment"
+    weight: 0.35
+    description: "Sustainable coordination framework"

package/courses/competitive-battlecards/scenarios/level-3/displacement-battlecards.yaml

@@ -0,0 +1,69 @@
+meta:
+  id: displacement-battlecards
+  level: 3
+  course: competitive-battlecards
+  type: output
+  description: "Build displacement battlecards — create specialized cards for ripping out incumbents that address switching costs, migration, and change management"
+  tags: [sales, battlecards, displacement, switching, migration, incumbent, advanced]
+
+state: {}
+
+trigger: |
+  You're the Head of Competitive Strategy at "NextGen ERP" (cloud
+  ERP platform, $120M ARR). 80% of your target market is running
+  SAP or Oracle ERP on-premise. Your product is cloud-native,
+  mid-market focused, and 60% less expensive. But displacement deals
+  are your hardest — incumbents have deep roots.
+
+  The displacement challenge:
+  - SAP/Oracle have been installed for 10-20 years
+  - Custom integrations, workflows, and reports built on top
+  - Thousands of users trained on the old system
+  - IT teams whose jobs are maintaining the incumbent
+  - Sunk cost psychology: "We've invested $5M in SAP, we can't
+    walk away"
+  - Risk aversion: "Nobody ever got fired for buying SAP"
+
+  Your displacement success data:
+  - Win rate in greenfield (no incumbent): 45%
+  - Win rate displacing SAP: 12%
+  - Win rate displacing Oracle: 18%
+  - Average displacement cycle: 14 months (vs. 5 months greenfield)
+  - Displaced customers report 40% cost reduction after migration
+
+  Tasks:
+  1. Create the "Displacement Battlecard for SAP" — this is NOT a
+     standard competitive battlecard. It must address: the emotional
+     barriers (sunk cost, fear of change), the practical barriers
+     (migration cost, data transfer, user retraining), the political
+     barriers (SAP consultants and IT team resistance), and the
+     financial case (TCO comparison including migration costs). Each
+     section needs specific content, not just categories.
+
+  2. Write the "Displacement Deal Qualification Criteria" — not
+     every SAP/Oracle customer is ready to switch. Create the
+     scoring framework (10 criteria) that helps reps identify
+     which accounts are displacement-ready (high pain, contract
+     expiring, new leadership) vs. which are too deeply entrenched.
+     Include the score threshold for "pursue" vs. "nurture" vs.
+     "avoid."
+
+  3. Design the "Phased Displacement Strategy" — for accounts scored
+     as displacement-ready, map the battlecard content to a 14-month
+     deal cycle. Show which battlecard content is used in each phase:
+     awareness (months 1-3), evaluation (months 4-7), decision
+     (months 8-11), and migration commitment (months 12-14).
+
+assertions:
+  - type: llm_judge
+    criteria: "Displacement battlecard addresses all barrier types — emotional (sunk cost, fear), practical (migration, retraining), political (IT resistance, consultant influence), and financial (TCO with migration costs included) with specific content for each"
+    weight: 0.35
+    description: "Complete displacement battlecard"
+  - type: llm_judge
+    criteria: "Qualification criteria are discriminating — 10 criteria that genuinely separate displacement-ready accounts from entrenched ones, with a clear scoring system and threshold that prevents reps from wasting time on unwinnable deals"
+    weight: 0.35
+    description: "Discriminating qualification criteria"
+  - type: llm_judge
+    criteria: "Phased strategy maps content to timeline — shows which battlecard content is relevant at each stage of a 14-month cycle, reflecting the reality that displacement messaging evolves significantly from awareness to migration commitment"
+    weight: 0.30
+    description: "Timeline-mapped phased strategy"

package/courses/competitive-battlecards/scenarios/level-3/ethical-competitive-intelligence.yaml

@@ -0,0 +1,82 @@
+meta:
+  id: ethical-competitive-intelligence
+  level: 3
+  course: competitive-battlecards
+  type: output
+  description: "Navigate ethical boundaries in competitive intelligence — understand what's acceptable, what crosses lines, and how to build a CI program with integrity"
+  tags: [sales, battlecards, ethics, competitive-intelligence, integrity, legal, advanced]
+
+state: {}
+
+trigger: |
+  You're the VP of Marketing at "SecureNet" (cybersecurity SaaS,
+  $75M ARR). Your Competitive Intelligence team has brought you
+  5 intelligence-gathering scenarios and wants your ruling on
+  whether each is ethical, unethical, or in a gray area.
+
+  Scenario 1: Mystery shopping
+  Your CI analyst created a fake email address and signed up for a
+  CrowdStrike demo pretending to be a prospect. They recorded the
+  demo and extracted pricing, positioning, and product capabilities.
+  The CI analyst's reasoning: "This is public information — anyone
+  can sign up for a demo."
+
+  Scenario 2: Social media intelligence
+  Your team found a competitor's VP of Product posting about their
+  upcoming product roadmap on LinkedIn. They screenshotted it and
+  shared it internally. A week later, the VP deleted the post.
+  CI analyst's reasoning: "It was public. We didn't hack anything."
+
+  Scenario 3: Former competitor employee
+  You just hired an engineer who worked at Palo Alto for 3 years.
+  Your CI team wants to interview her about Palo Alto's architecture,
+  product roadmap, and internal challenges.
+  CI analyst's reasoning: "She's our employee now. Her knowledge is
+  her knowledge."
+
+  Scenario 4: Customer reference manipulation
+  Your CI team found that a key Fortinet customer (listed as a
+  case study on Fortinet's website) is unhappy and considering
+  leaving. Your team wants to contact them directly to gather
+  negative intelligence about Fortinet.
+  CI analyst's reasoning: "They're a prospect for us too. We're
+  doing outreach."
+
+  Scenario 5: Competitive bidding intelligence
+  A prospect accidentally included you on an email thread that
+  contained CrowdStrike's pricing proposal. You now know exactly
+  what CrowdStrike bid.
+  CI analyst's reasoning: "They sent it to us. We didn't steal it."
+
+  Tasks:
+  1. For each scenario, provide the ethics ruling: ethical, unethical,
+     or gray area. For each, explain: the principle at stake, the
+     risk if caught, the industry standard (SCIP Code of Ethics),
+     and your recommended action. Be specific — "it depends" is not
+     an answer.
+
+  2. Write the "Competitive Intelligence Ethics Policy" for SecureNet.
+     Include: the 5 principles of ethical CI, the specific practices
+     that are approved (with examples), the specific practices that
+     are prohibited (with examples), the gray areas that require
+     management approval, and the consequences for violations.
+
+  3. Your CI analyst argues: "Our competitors don't play fair. Why
+     should we?" Write the business case for ethical CI — why
+     integrity is a competitive advantage, not a disadvantage.
+     Include: reputational risk, legal exposure, customer trust
+     impact, and talent attraction.
+
+assertions:
+  - type: llm_judge
+    criteria: "Ethics rulings are well-reasoned — each scenario gets a clear ruling with principle-based reasoning, not just opinion, referencing SCIP ethics standards and distinguishing between public information, deception, trade secrets, and accidental disclosure"
+    weight: 0.35
+    description: "Well-reasoned ethics rulings"
+  - type: llm_judge
+    criteria: "Ethics policy is comprehensive and specific — 5 clear principles with specific approved and prohibited practices, gray area guidance, and enforcement consequences, usable as an actual company policy"
+    weight: 0.35
+    description: "Comprehensive ethics policy"
+  - type: llm_judge
+    criteria: "Business case for ethical CI is convincing — explains why integrity is strategically advantageous (reputation, legal protection, customer trust, talent) rather than just morally right, making the case in business language"
+    weight: 0.30
+    description: "Convincing ethics business case"

package/courses/competitive-battlecards/scenarios/level-4/ai-powered-battlecards.yaml

@@ -0,0 +1,69 @@
+meta:
+  id: ai-powered-battlecards
+  level: 4
+  course: competitive-battlecards
+  type: output
+  description: "Leverage AI for competitive intelligence and battlecard creation — use AI tools to monitor competitors, generate content, and personalize battlecards at scale"
+  tags: [sales, battlecards, AI, automation, intelligence, personalization, expert]
+
+state: {}
+
+trigger: |
+  You're the Director of Revenue Operations at "InsightPro"
+  (revenue intelligence platform, $80M ARR). Your CEO wants to
+  "AI-enable" the competitive intelligence program. Currently,
+  your CI analyst spends 30 hours/week manually monitoring
+  competitors and updating battlecards. The CEO says: "AI should
+  do 80% of this."
+
+  Current CI workflow (manual):
+  - 15 hours/week: Monitoring competitor websites, news, social
+  - 8 hours/week: Analyzing changes and writing FIA briefs
+  - 5 hours/week: Updating battlecard content
+  - 2 hours/week: Responding to field requests
+
+  AI capabilities available:
+  - LLM-powered content generation (GPT-4, Claude)
+  - Competitive intelligence platforms with AI (Crayon, Klue)
+  - Web scraping and monitoring tools
+  - Natural language processing for review/sentiment analysis
+  - CRM integration APIs
+
+  Your budget: $150K for AI-powered CI tools (annual)
+
+  Tasks:
+  1. Design the "AI-Powered CI Workflow" — map how AI transforms
+     each of the 4 current workflow components. For each: what AI
+     handles (automated monitoring, draft generation, pattern
+     detection), what humans still handle (validation, strategy,
+     nuanced analysis), and the expected time savings. Be realistic —
+     AI hallucination risk is real in competitive content.
+
+  2. Write the "AI Battlecard Generation" specification — how you'd
+     use AI to auto-generate battlecard drafts from competitive
+     intelligence signals. Include: the prompt engineering approach,
+     the human review process (AI drafts, human validates), the
+     quality gates (what must be human-verified before publishing),
+     and the risk mitigation for AI-generated competitive claims.
+
+  3. Design the "AI Personalization Engine" — a system that
+     generates deal-specific battlecard content in real-time. When a
+     rep enters a deal (competitor X, industry Y, persona Z), the
+     system generates a personalized battlecard. Include: the data
+     inputs, the generation logic, the accuracy validation, and the
+     UX for reps. Address the risk: what happens when AI generates
+     an incorrect competitive claim and a rep uses it?
+
+assertions:
+  - type: llm_judge
+    criteria: "AI workflow is realistic about capabilities — clearly separates what AI can do (monitoring, draft generation, pattern detection) from what requires human judgment (validation, strategy, nuance), with honest time savings expectations"
+    weight: 0.35
+    description: "Realistic AI workflow"
+  - type: llm_judge
+    criteria: "AI generation spec addresses hallucination risk — includes prompt engineering, mandatory human review, quality gates for competitive claims, and specific risk mitigation for AI-generated content that could be factually wrong"
+    weight: 0.35
+    description: "Hallucination-aware AI generation"
+  - type: llm_judge
+    criteria: "Personalization engine is practical — describes inputs, generation logic, and validation that would work at deal-level personalization, with clear risk mitigation for incorrect AI claims"
+    weight: 0.30
+    description: "Practical personalization engine"