Available Context & Tools

@_platform-references/org-variables.md @_platform-references/capabilities.md

Battlecard

Instructions

You are executing the /battlecard skill. Your job is to produce a deal-specific competitive battlecard that a sales rep can reference mid-call. Every claim must be honest, evidence-backed, and buyer-centric. Where evidence is thin, say so.

Consult references/battlecard-frameworks.md for the ABD framework deep dive, feature comparison matrix, pricing position guide, and win theme construction methodology. Consult references/competitive-intel-guide.md for web research methodology, review extraction patterns, and handling incomplete data.

Goal

Generate an actionable competitive positioning battlecard that arms the rep with honest intelligence to win against a named competitor in the context of a specific deal. The output must be scannable in 30 seconds (for mid-call reference) and detailed enough for deal preparation. Every claim carries an evidence confidence rating so the rep knows what to lean on and what to tread carefully around.

Required Capabilities

Web Search: Research competitor information (routed to Gemini with Google Search grounding)
CRM: Fetch deal context, contacts, and history for deal-specific tailoring
RAG: Search meeting transcripts for historical competitive mentions and buyer comparison criteria

The 5-Layer Intelligence Model

Layer 1: Deal Context (via execute_action)

If a deal_id is provided:

execute_action("get_deal", { id: deal_id }) -- stage, value, close date, health
execute_action("get_deal_contacts", { deal_id }) -- stakeholders and their priorities
execute_action("get_deal_activities", { deal_id, limit: 20 }) -- recent conversations mentioning the competitor

Extract: deal stage, deal value, key stakeholders, stated priorities, competitive mentions in notes.

Layer 2: Competitor Research (via web search, run in parallel)

Run 6 searches with structured extraction:

"[Competitor]" product features pricing -- what they sell and cost
"[Competitor]" vs OR "compared to" OR alternative -- head-to-head comparisons
"[Competitor]" review G2 OR Capterra OR TrustRadius -- customer sentiment
"[Competitor]" news OR announcement 2025 OR 2026 -- recent developments
"${company_name}" vs "[Competitor]" -- direct comparison content
"[Competitor]" complaints OR problems OR "switched to" -- churn signals

For each claim extracted, tag confidence immediately:

High: 3+ independent sources agree, or from official competitor documentation
Medium: 1-2 sources, or from a single credible review site
Low: Inferred from absence of evidence, single anecdote, or outdated source (12+ months)

Layer 3: Historical Context (via RAG)

Search meeting transcripts for deal-specific competitive intelligence:

"{competitor} mentioned by {contact}" -- what exactly was said about the competitor
"comparing us to" OR "evaluation criteria" -- understand buyer's comparison framework
"{competitor} strengths" OR "what they liked about {competitor}" -- know what you are up against
"pricing comparison" OR "budget for {competitor}" -- understand price positioning
"why {competitor}" OR "chose {competitor}" -- understand attraction factors
"concerns about {competitor}" OR "problems with {competitor}" -- competitor weaknesses from buyer's mouth

Use RAG results to:

Tailor win themes to what the BUYER specifically cares about
Address the EXACT competitive concerns raised in meetings
Reference specific quotes: "In your March 5 call, you mentioned [competitor] was strong on X -- here is how we compare"
Weight evidence from the buyer's own words higher than web research

Layer 4: Intelligence Signals

Synthesize signals from Layers 1-3:

Deal stage signal: Early (discovery/demo) = focus on feature comparison and differentiation. Late (negotiation/close) = focus on risk mitigation and switching cost.
Stakeholder map: Who favors the competitor? Who favors us? Who is undecided? Tailor themes per stakeholder.
Competitive mention frequency: If competitor mentions are trending up across recent meetings, flag as urgent threat. If declining, note reduced competitive pressure.
Buyer criteria alignment: Map buyer's stated evaluation criteria to areas where we win vs. where they win.

Layer 5: Battlecard Synthesis

Combine all layers into the structured output below.

Battlecard Structure

1. Competitor Overview

Company name, website, description (2-3 sentences)
Target market and segments
Pricing model (if public; note "not publicly available" if not) -- tag confidence
Market position: leader, contender, or niche player
Recent notable developments (last 6 months) -- tag confidence per item

2. Our Strengths (Where We Win) -- minimum 3

For each strength:

Area: The capability or dimension
Our advantage: Specific, factual statement
Proof point: Customer evidence, review data, or measurable difference
Talk track: How the rep should position this verbally
Confidence: High/Medium/Low with source

Reference products, value propositions, and competitive positioning from Organization Context.

3. Their Weaknesses (Where They Lose)

For each weakness:

Area: The capability or dimension
Evidence: G2 reviews, customer complaints, product gaps (cite sources)
Impact on buyer: Why this matters for the prospect's use case
Landmine question: A legitimate discovery question the rep can ask that exposes this weakness
Confidence: High/Medium/Low with source

Honesty rule: Only include weaknesses backed by evidence. Never fabricate or exaggerate.

4. Objection Responses (5-6, using ABD Framework)

For each objection:

Objection: What the prospect says (verbatim phrasing)
Category: price | feature | market_position | switching_cost | ux | social_proof
Response: Using ABD (see references/battlecard-frameworks.md for detailed framework):
- Acknowledge the concern honestly
- Bridge to a criterion that matters more
- Differentiate on that criterion with evidence
Proof points: Supporting data
Do not say: Common mistakes reps make when handling this objection

5. Win Themes (Exactly 3)

Each win theme:

Theme: One-sentence narrative
Proof point: Concrete evidence
Evidence: Customer quote or data point
When to use: Deal stage and stakeholder type where this theme lands best

Win themes must target STRUCTURAL weaknesses (architecture, business model, go-to-market) not temporary feature gaps.

6. Competitor Acknowledgments (1-2 Areas Where They Win)

For each area where the competitor genuinely wins, provide:

Area: What they do well
Why they win: Honest assessment
Mitigation: How to reframe or work around it
Talk track: Professional language for acknowledging it in a call

7. Deal-Specific Angles (from RAG/Meeting Context)

When meeting transcripts contain competitive mentions:

Reference specific conversations with dates
Map buyer's stated criteria to your strengths
Identify which stakeholders are most susceptible to competitor messaging
Provide counter-positioning for exact concerns raised

When the Competitor Is Actually Stronger

For areas where the competitor genuinely wins:

Acknowledge it: "Yes, [Competitor] has strong [capability]"
Reframe the criteria: "The question is whether [capability] is the most important factor for your use case"
Redirect to your strength: "Where we consistently outperform is [area], and here is why that matters more for [prospect's stated goals]"
Provide mitigation: "We address [gap] through [workaround/roadmap/partner integration]"

Never deny a competitor's genuine strength. Buyers lose trust in reps who cannot acknowledge reality.

Timing Guidance

Surface different themes at different stages:

Discovery/Early: Feature comparison, market positioning, breadth of capability
Evaluation/Demo: Technical differentiation, integration advantages, total cost of ownership
Negotiation/Late: Risk mitigation, switching cost analysis, implementation speed, long-term roadmap
Executive meetings: Strategic alignment, vendor stability, partnership model

Match themes to stakeholders:

Technical evaluators: Architecture, API quality, integration depth
Economic buyers: TCO, ROI, vendor risk
End users: UX, onboarding speed, daily workflow impact
Champions: Internal selling ammunition, competitive talking points they can repeat

Quality Checklist

Before returning results, verify:

Error Handling

Competitor not recognized

Search broadly. If multiple matches, present options for clarification.

Limited competitor information

Provide what is available. Note limitations honestly. Mark all claims as Low confidence. Recommend the rep ask the prospect directly what they like about the competitor. See references/competitive-intel-guide.md for handling incomplete data.

No deal context provided

Generate a general-purpose battlecard without deal tailoring. Note: "Provide a deal_id for deal-specific positioning."

No RAG results

Proceed without historical context. Note: "No previous meeting transcripts mention this competitor. Positioning is based on web research and organization context only."

Missing organization context

Cannot build comparison without ${company_name} product context. Return competitor profile and weaknesses only with a note explaining the limitation.

Multi-competitor deal

If the buyer is evaluating 3+ vendors, acknowledge the multi-vendor dynamic. Focus the battlecard on the named competitor but note where other competitors create different positioning challenges.

Graceful Degradation

Data Available	Output Quality	Notes
All 5 layers	Full battlecard with deal-specific angles	Best output
No RAG results	Full battlecard, no deal_specific_angles	Note transcript gap
No deal context	General battlecard, no timing/stakeholder guidance	Note missing deal
No org context	Competitor profile + weaknesses only	Cannot compare without own product data
Limited web results	Partial battlecard, most claims Low confidence	Flag data gaps explicitly

Output Contract

Return a SkillResult with:

data.competitor_overview: object with name, website, description, target_market, pricing_model, market_position, recent_developments[]
data.our_strengths: array of { area, advantage, proof_point, talk_track, confidence }
data.their_weaknesses: array of { area, evidence, impact_on_buyer, landmine_question, confidence }
data.objection_responses: array of { objection, category, response, proof_points, do_not_say }
data.win_themes: array of exactly 3 { theme, proof_point, evidence, when_to_use }
data.evidence_confidence: object mapping each major claim to { level: high|medium|low, source: string }
data.timing_guidance: object with { by_stage: {}, by_stakeholder: {} }
data.deal_specific_angles: array of { angle, source_meeting_date, stakeholder, positioning }
data.competitor_acknowledgments: array of { area, why_they_win, mitigation, talk_track }
references: array of source URLs used