train

/octave:train - Sales Training Ground

Practice and learn your GTM playbook through interactive role-play simulations and knowledge quizzes — all grounded in your real library data. Role-play against buyer personas with realistic objections, or quiz yourself on value props, competitive positioning, and discovery techniques.

Usage

/octave:train [mode] [--persona <name>] [--competitor <name>] [--difficulty easy|medium|hard]

Modes

/octave:train                                          # Interactive - pick a mode
/octave:train roleplay                                 # Simulate a buyer conversation
/octave:train roleplay --persona "CTO"                 # Role-play with a specific persona
/octave:train roleplay --scenario discovery            # Practice discovery calls
/octave:train quiz                                     # Test your GTM knowledge
/octave:train quiz --topic objections                  # Quiz on objection handling
/octave:train quiz --competitor "Acme"                 # Competitive knowledge check

Instructions

When the user runs /octave:train:

Step 1: Choose Mode

If no mode specified, ask:

AskUserQuestion({
  questions: [{
    question: "What do you want to practice?",
    header: "Train mode",
    options: [
      { label: "Role-Play", description: "Simulate a sales conversation — I'll play the buyer and give you feedback" },
      { label: "Quiz", description: "Test your knowledge of personas, objections, value props, and competitive positioning" },
      { label: "Guided Learning", description: "Walk me through a topic from your playbook — teach me like I'm a new hire" }
    ],
    multiSelect: false
  }]
})

---

Mode: Role-Play

Simulate realistic buyer conversations using persona data from the library.

Step RP-1: Setup the Scenario

Ask for scenario parameters (use AskUserQuestion for structured choices):

AskUserQuestion({
  questions: [
    {
      question: "What scenario do you want to practice?",
      header: "Scenario",
      options: [
        { label: "Discovery call", description: "First conversation — qualify the opportunity and uncover pain" },
        { label: "Objection handling", description: "Practice responding to tough objections mid-deal" },
        { label: "Demo pitch", description: "Present your product's value to a skeptical buyer" },
        { label: "Competitive displacement", description: "Sell against a competitor the buyer currently uses" }
      ],
      multiSelect: false
    },
    {
      question: "How tough should I be?",
      header: "Difficulty",
      options: [
        { label: "Friendly", description: "Interested buyer, open to learning — good for building confidence" },
        { label: "Skeptical (Recommended)", description: "Realistic buyer who pushes back and needs convincing" },
        { label: "Hostile", description: "Tough buyer — time-pressed, has objections, hard to impress" }
      ],
      multiSelect: false
    }
  ]
})

If no persona specified, present available personas:

# Get personas from library
list_all_entities({ entityType: "persona" })

Ask:

AskUserQuestion({
  questions: [{
    question: "Which buyer persona should I play?",
    header: "Persona",
    options: [
      { label: "[Persona 1 name]", description: "[Title] — [key concern]" },
      { label: "[Persona 2 name]", description: "[Title] — [key concern]" },
      { label: "[Persona 3 name]", description: "[Title] — [key concern]" }
    ],
    multiSelect: false
  }]
})

Step RP-2: Load Persona Intelligence

# Get full persona details
get_entity({ oId: "<persona_oId>" })

# Get matching playbook with value props
search_knowledge_base({ query: "<persona name> <scenario>", entityTypes: ["playbook"] })
get_playbook({ oId: "<playbook_oId>", includeValueProps: true })

# Get real objections from conversations (to make role-play realistic)
list_findings({
  query: "objections pushback concerns",
  startDate: "<180 days ago>",
  eventFilters: {
    personas: ["<persona_oId>"]
  }
})

# Get product details
list_all_entities({ entityType: "product" })
get_entity({ oId: "<product_oId>" })

# Get competitor details (for competitive scenarios)
get_entity({ oId: "<competitor_oId>" })  // if competitive scenario

# Get proof points (to evaluate if rep uses them)
search_knowledge_base({ query: "<persona> results metrics", entityTypes: ["proof_point", "reference"] })

Step RP-3: Set the Scene

Present the scenario context, then begin:

See roleplay-scene-template.md for the role-play scene template.

How to play the buyer:

Use the loaded persona data to respond realistically:

• Reference real pain points from the persona entity
• Raise real objections from conversation findings data
• React based on difficulty level:
  • Friendly: Engaged, asks questions, shares information willingly
  • Skeptical: Needs proof, challenges claims, asks "why should I care?"
  • Hostile: Short answers, pushes on price, questions ROI, brings up competitors
• Respond naturally to what the user says — don't just cycle through objections
• If the user makes a strong point, acknowledge it (even skeptical buyers respond to good selling)
• If the user fumbles, the buyer gets more distant/disengaged

End the conversation after 8-12 exchanges with a natural conclusion:

• Friendly: "This sounds interesting, let's set up a follow-up"
• Skeptical: "I need to think about it" or "Send me some materials"
• Hostile: "I don't think this is for us" (unless the user sold well)

Step RP-4: Scorecard

After the role-play ends, provide detailed feedback:

See roleplay-scorecard-template.md for the role-play scorecard template.

---

Mode: Quiz

Test knowledge of the user's own GTM library.

Step Q-1: Choose Topic

AskUserQuestion({
  questions: [{
    question: "What do you want to be quizzed on?",
    header: "Quiz topic",
    options: [
      { label: "Personas", description: "Pain points, priorities, buying triggers, and how to sell to each persona" },
      { label: "Objection handling", description: "Common objections and how to respond — from your playbook and real conversations" },
      { label: "Competitive positioning", description: "Differentiators, trap questions, and counters for each competitor" },
      { label: "Full GTM review", description: "Mix of everything — personas, products, value props, objections, competitors" }
    ],
    multiSelect: false
  }]
})

Also ask for quiz format:

AskUserQuestion({
  questions: [{
    question: "What format?",
    header: "Format",
    options: [
      { label: "Rapid fire (Recommended)", description: "Quick question-answer, 10 questions, scored at the end" },
      { label: "Scenario-based", description: "Situational questions — 'A prospect says X, what do you do?'" },
      { label: "Deep dive", description: "Fewer questions but explain your reasoning — I'll coach you on each answer" }
    ],
    multiSelect: false
  }]
})

Step Q-2: Load Quiz Material

# Load based on topic
# For Personas:
list_all_entities({ entityType: "persona" })
get_entity({ oId: "<persona_oId>" })  // for each persona

# For Objections:
list_findings({
  query: "objections pushback concerns pricing",
  startDate: "<180 days ago>"
})
get_playbook({ oId: "<playbook_oId>", includeValueProps: true })

# For Competitive:
list_all_entities({ entityType: "competitor" })
get_entity({ oId: "<competitor_oId>" })  // for each competitor

# For Full GTM:
list_all_entities({ entityType: "persona" })
list_all_entities({ entityType: "product" })
list_all_entities({ entityType: "competitor" })
search_knowledge_base({ query: "value propositions proof points" })
list_all_entities({ entityType: "use_case" })

Step Q-3: Run the Quiz

See quiz-formats.md for the rapid fire, scenario-based, and deep dive quiz format templates with question types per topic.

Step Q-4: Quiz Results

See quiz-results-template.md for the quiz results template.

---

Mode: Guided Learning

Walk through a topic from the library like a training session.

Step GL-1: Choose Topic

AskUserQuestion({
  questions: [{
    question: "What do you want to learn about?",
    header: "Topic",
    options: [
      { label: "A persona", description: "Deep walkthrough of how to sell to a specific buyer type" },
      { label: "A competitor", description: "Learn competitive positioning, differentiators, and counters" },
      { label: "A playbook", description: "Walk through a playbook's strategy, value props, and objection handling" },
      { label: "Your product", description: "Master your product's capabilities, use cases, and proof points" }
    ],
    multiSelect: false
  }]
})

Step GL-2: Load and Teach

Fetch the relevant entity and present it as a structured training walkthrough:

# Load the entity
get_entity({ oId: "<entity_oId>" })

# Load related playbook
get_playbook({ oId: "<playbook_oId>", includeValueProps: true })

# Load real conversation examples
list_findings({
  query: "<topic>",
  startDate: "<180 days ago>"
})

# Load proof points
search_knowledge_base({ query: "<topic>", entityTypes: ["proof_point", "reference"] })

See guided-learning-template.md for the interactive guided learning lesson template.

MCP Tools Used

Library Context

• list_all_entities - List personas, products, competitors, use cases
• get_entity - Full entity details (persona pain points, competitor weaknesses, etc.)
• get_playbook - Playbook with value props, discovery questions, objection handling
• list_value_props - Value propositions by persona
• search_knowledge_base - Proof points, references, messaging

Conversation Evidence

• list_findings - Real objections, pain points, and signals from calls/emails
• list_events - Deal outcomes (win/loss evidence for coaching)
• get_event_detail - Specific interaction details for training examples

Content Generation

• generate_content - Generate scenario descriptions, coaching feedback

Error Handling

No Personas in Library:

No personas found in your library.

Role-play and quizzes work best with persona data. Add personas first: /octave:library create persona

I can still run a general sales quiz using your product info.

No Conversation Data:

No conversation data available yet.

I'll use your library data for role-play and quizzes. As your team logs calls and emails, training will get richer with real-world objections and patterns.

No Competitors:

No competitors in your library.

Competitive quizzes and displacement role-plays need competitor data. Add competitors: /octave:library create competitor

I can still quiz you on personas, value props, and general objection handling.

Sparse Library:

Your library has limited data for a full training session.

Start with:

1. /octave:library create product - Add your product
2. /octave:library create persona - Add buyer personas
3. /octave:library create competitor - Add competitors

Even with just a product and one persona, I can run basic training.

Related Skills

• /octave:enablement - Generate training materials (cheat sheets, objection guides, discovery banks)
• /octave:battlecard - Deep competitive intelligence for competitive training
• /octave:insights - Surface real field intelligence to inform training
• /octave:wins-losses - Win/loss patterns to learn from
• /octave:research - Research a real prospect before a live call
• /octave:generate - Generate real outreach after practicing