M&A Toolkit

Unified interface for all Nuvini M&A operations. This skill consolidates 6 specialized M&A capabilities into a single, streamlined workflow.

Quick Reference

Workflow Overview

┌─────────────────────────────────────────────────────────────────┐
│                    M&A DEAL PIPELINE                             │
├─────────────────────────────────────────────────────────────────┤
│                                                                  │
│  1. SCREENING          2. ANALYSIS           3. APPROVAL        │
│  ┌──────────────┐     ┌──────────────┐     ┌──────────────┐    │
│  │ /mna triage  │ ──► │ /mna extract │ ──► │ /mna proposal│    │
│  │ Score: 0-10  │     │ CIM data     │     │ IRR/MOIC     │    │
│  └──────────────┘     └──────────────┘     └──────────────┘    │
│        │                    │                    │              │
│        │ Score ≥ 7         │                    │ IRR ≥ 20%    │
│        ▼                    ▼                    ▼              │
│  ┌──────────────┐     ┌──────────────┐     ┌──────────────┐    │
│  │ /mna aimpact │     │ /mna analysis│     │ /mna deck    │    │
│  │ AI savings   │     │ 18-page PDF  │     │ Board PPT    │    │
│  └──────────────┘     └──────────────┘     └──────────────┘    │
│                                                                  │
└─────────────────────────────────────────────────────────────────┘

1. Triage Analysis (/mna triage)

Score M&A opportunities 0-10 against Nuvini investment criteria.

Required Inputs

• •Company name
• •Revenue (annual, millions BRL)
• •EBITDA (millions BRL)

Optional Inputs

• •Revenue growth rate (%)
• •EBITDA margin (%)
• •Recurring revenue %
• •Churn rate (%)
• •Customer count
• •Business model (SaaS, Software, B2B)
• •Market vertical

Scoring Framework

Red Flags (Auto-Reject)

• •Not SaaS/Software business
• •Not Brazil-focused
• •EBITDA <R$5M or >R$100M
• •Churn >15%
• •Cash conversion <60%

Example

User: /mna triage
Company: TechBrasil
Revenue: R$50M, EBITDA: R$15M
Growth: 25%, SaaS, tax tech vertical

Output: Score 8.2/10 - PROCEED
- Strengths: Strong margins, vertical focus
- Risks: Customer concentration
- Next: Generate proposal

2. Financial Data Extraction (/mna extract)

Extract structured data from CIMs, financial statements, and Excel models.

Supported Formats

• •PDF: CIMs, financial statements, pitch decks, reports
• •Excel: Financial models, data exports, projections

Extraction Modes

• •figures - Key financial metrics (revenue, EBITDA, etc.)
• •tables - All tables with page numbers
• •all - Complete extraction

Model Selection

• •Sonnet 4.5: Standard extraction (95% of tasks)
• •Opus 4.1: High-stakes validation (>$25M deals, ambiguous docs)

Example

User: /mna extract ProjectAlpha_CIM.pdf

Output:
{
  "company": "Project Alpha",
  "revenue_2024": 42000000,
  "ebitda_2024": 12600000,
  "ebitda_margin": 0.30,
  "recurring_revenue_pct": 0.92,
  "employees": 85,
  "customers": 450
}

3. Proposal Generation (/mna proposal)

Generate comprehensive financial models with IRR/MOIC calculations.

Required Inputs

• •Company name
• •EBITDA by year (historical + projected)

Default Deal Structure

• •Purchase: 6.0x EBITDA
• •Cash at close: 60%
• •Deferred: 40% in Year 1
• •Earnout: 3.0x on EBITDA growth
• •Debt: 6-year bullet at 9% PIK

Outputs

• •IRR and MOIC calculations
• •Payment schedule (target perspective)
• •Sources & uses (acquirer perspective)
• •Debt schedule
• •Sensitivity analysis
• •Excel model with formulas

Hurdle Rates

• •Minimum IRR: 20%
• •Minimum MOIC: 2.5x
• •Maximum leverage: 4.0x EBITDA

Example

User: /mna proposal TechBrasil
EBITDA: 2024: R$15M, 2025: R$18M, 2026: R$22M

Output:
- IRR: 35.2%
- MOIC: 4.2x
- Purchase: R$90M (6.0x)
- Excel: TechBrasil_proposal.xlsx

4. Investment Analysis (/mna analysis)

Generate professional 18-page investment analysis PDF.

Page Structure

Data Sources

• •Triage analysis results
• •CIM extraction data
• •AI-synthesized insights

Example

User: /mna analysis TechBrasil (triage_id: abc123)

Output:
- File: investment_analysis_techbrasil_20250110.pdf
- Pages: 18
- Extraction quality: high

5. Board Presentation (/mna deck)

Create PowerPoint presentations for investment committee.

Formats

Summary (5 slides)

1. •Executive Dashboard
2. •Financial Highlights
3. •Risk Assessment
4. •Recommendations
5. •Next Steps

Full Analysis (20+ slides) All summary slides plus:

• •Deal structure details
• •Financial deep-dive
• •Sensitivity analysis
• •Integration framework
• •Exit strategy
• •Implementation roadmap

Brand Standards

• •Primary Blue: #1B4F8C
• •Accent Teal: #00A19C
• •Highlight Orange: #FF8C42
• •Font: Montserrat (headers), Open Sans (body)
• •Quality: McKinsey/BCG standard

Example

User: /mna deck TechBrasil
IRR: 35%, MOIC: 4.2x, Investment: R$90M

Output:
- File: TechBrasil_board_deck.pptx
- Slides: 5 (summary format)
- Recommendation: STRONGLY APPROVE

6. AI Impact Analysis (/mna aimpact)

Analyze AI cost reduction opportunities for acquisition targets.

Purpose

Model potential cost savings from AI implementation post-acquisition.

Analysis Categories

Three Scenarios

• •Conservative: 8-18% reduction, <12mo payback
• •Moderate: 15-30% reduction, <8mo payback
• •Aggressive: 25-45% reduction, <6mo payback

Outputs

• •Excel model with NPV calculations
• •Word document with recommendations
• •Implementation roadmap (12 months)

Example

User: /mna aimpact TechBrasil financials.xlsx

Output:
- Excel: TechBrasil_ai_cost_reduction.xlsx
- Word: TechBrasil_ai_recommendations.docx
- Conservative savings: R$2.1M/year
- Moderate savings: R$3.8M/year
- Aggressive savings: R$5.5M/year

Atomic Tools vs Outcomes

Agent-Native Design Principle

This skill follows the Agent-native principle: "Tools should be atomic primitives. Features are outcomes achieved by an agent operating in a loop."

Atomic Tools Used

The mna-toolkit uses these atomic primitives:

• •Read - Read CIMs, Excel files, previous analyses
• •Write - Generate reports, proposals, presentations
• •Edit - Modify financial models, update documents
• •Glob/Grep - Find deal files, search for financial data
• •Bash - Execute Python scripts, generate PDFs, run calculations
• •WebFetch/WebSearch - Research companies, gather market data
• •TaskCreate/TaskUpdate/TaskList - Track multi-step deal workflows

Key insight: The six sub-commands (/mna triage, /mna extract, etc.) are entry points that describe desired outcomes, not hardcoded workflow functions.

Outcomes Achieved via Prompts

Each sub-command is an outcome:

1. •

   /mna triage - Outcome: "Score this deal 0-10 against Nuvini criteria and identify red flags"

   • •Agent uses Read to get deal data, applies scoring criteria via prompts, uses Write to create JSON report
   • •Scoring criteria are in prompts, not hardcoded logic
2. •

   /mna extract - Outcome: "Extract structured financial data from this document"

   • •Agent uses Read to analyze PDF/Excel, follows extraction prompts, uses Write to create JSON/CSV
   • •Extraction patterns are prompt-based, allowing flexibility for different document formats
3. •

   /mna proposal - Outcome: "Generate IRR/MOIC financial model with Nuvini deal structure"

   • •Agent calculates via Bash (Python scripts), follows Excel template prompts, uses Write for output
   • •Deal structure parameters are in prompts, easily modified without code changes
4. •

   /mna analysis - Outcome: "Create professional 18-page investment analysis PDF"

   • •Agent synthesizes data from multiple sources, follows page-by-page prompts, uses Bash to generate PDF
   • •Page structure is prompt-defined, not template-locked
5. •

   /mna deck - Outcome: "Create board-quality PowerPoint with key metrics"

   • •Agent selects relevant data, follows slide design prompts, uses Bash (python-pptx) to generate
   • •Slide content adapts based on deal specifics, guided by prompts
6. •

   /mna aimpact - Outcome: "Model AI cost reduction scenarios for this target"

   • •Agent analyzes cost structure, applies AI opportunity prompts, calculates NPV via Bash
   • •Scenario assumptions are prompt-defined, easily adjusted for different industries

How to Modify Behavior via Prompts

Want to change triage scoring criteria?

Edit the scoring framework prompts (lines 92-98). Change the weights or add new categories:

### Scoring Framework (Modified for SaaS Focus)

| Category         | Max Points | Criteria                                       |
| ---------------- | ---------- | ---------------------------------------------- |
| Financial Health | 30         | EBITDA margin, revenue growth, cash conversion |
| Business Model   | 30         | Recurring revenue, churn rate, NRR             |
| Market Position  | 25         | Vertical focus, market share, leadership       |
| Tech Stack       | 15         | Modern architecture, API-first, AI-ready       |

The agent will adapt scoring based on these updated prompts.

Want different proposal assumptions?

Modify the default deal structure prompts (lines 172-177):

### Default Deal Structure (Modified)

- **Purchase**: 7.0x EBITDA (up from 6.0x for premium assets)
- **Cash at close**: 50% (more seller financing)
- **Deferred**: 30% in Year 1, 20% in Year 2
- **Earnout**: 2.5x on revenue growth (not EBITDA)
- **Debt**: 5-year amortizing at 8% (not PIK)

The agent's proposal calculations will reflect these new parameters.

Want to add new analysis pages?

Extend the page structure prompts (lines 216-235) with additional pages:

| 19   | AI Integration Opportunities |
| 20   | Synergies with Portfolio      |
| 21   | Go-to-Market Strategy         |

The agent will generate content for these new pages following the pattern.

Want different deck formats?

Create new slide structure prompts beyond "Summary (5 slides)" and "Full (20+ slides)":

### Formats

**Quick Pitch (3 slides)**
1. One-slide summary with key metrics
2. Investment thesis and deal structure
3. Risk/mitigation matrix

**Due Diligence Deep-Dive (40+ slides)**
- All full analysis slides
- Financial model walkthrough
- Customer concentration analysis
- Legal/regulatory review
- Integration planning timeline

Want industry-specific triage criteria?

Add conditional scoring based on vertical:

### Industry-Specific Red Flags

**Tax/Accounting Software**
- Regulatory changes risk
- Professional certification dependencies

**HR Tech**
- Data privacy (LGPD) compliance critical
- Customer contract length <2 years

**FinTech**
- Banking partnerships required
- Regulatory approval timelines

Contrast: Workflow-Shaped Anti-Pattern

What this skill does NOT do:

# Anti-pattern: Hardcoded workflow tool
def mna_complete_analysis(company_name, cim_file):
    # Hardcoded steps
    triage = run_triage_logic(company_name)
    if triage.score < 7:
        return "REJECTED"

    data = extract_with_predefined_fields(cim_file)
    proposal = calculate_with_fixed_multiples(data)
    analysis = generate_from_locked_template(data, proposal)
    deck = create_slides_from_template(analysis)

    return {triage, data, proposal, analysis, deck}

What this skill DOES:

Each sub-command is an outcome-oriented prompt:

• •"Evaluate this deal against Nuvini investment criteria and provide a scored recommendation"
• •"Extract relevant financial metrics from this document, adapting to its structure"
• •"Model the financial returns for this deal using Nuvini's typical structure"

The agent uses atomic tools (Read, Bash, Write) to achieve these outcomes. The workflow emerges from the agent's decision-making based on the document structure, financial data, and desired outcome.

Example: Evolving Triage Criteria

Scenario: Nuvini wants to prioritize AI-ready companies starting Q2 2026.

Without code changes, update the triage prompt:

### Scoring Framework (Updated Q2 2026 - AI Focus)

| Category         | Max Points | Criteria                                       |
| ---------------- | ---------- | ---------------------------------------------- |
| Financial Health | 30         | EBITDA margin, revenue growth, cash conversion |
| Business Model   | 25         | Recurring revenue, churn rate, customer base   |
| AI Readiness     | 25         | Data quality, API infra, AI use cases          |
| Strategic Fit    | 20         | Portfolio synergies, integration complexity    |

### AI Readiness Evaluation

Score 0-25 based on:
- Data infrastructure: APIs, clean datasets, real-time pipelines (10 pts)
- Current AI usage: ML models in production, AI features launched (8 pts)
- AI potential: Clear use cases, customer demand for AI features (7 pts)

The agent immediately adapts its triage scoring to include AI readiness assessment, asking relevant questions and evaluating companies differently, all without touching code.

Session Tracking

Use ${CLAUDE_SESSION_ID} to track analysis across sessions for comprehensive due diligence workflows.

Directory Structure

Organize multiple document extractions per deal with session-specific directories:

# Create session-specific output directories
DEAL_DIR="deals/${COMPANY_NAME}/sessions/${CLAUDE_SESSION_ID}"
mkdir -p "$DEAL_DIR"

# All outputs tagged with session ID
TRIAGE_FILE="$DEAL_DIR/triage_${CLAUDE_SESSION_ID}.json"
EXTRACTION_FILE="$DEAL_DIR/extraction_${CLAUDE_SESSION_ID}.json"
PROPOSAL_FILE="$DEAL_DIR/proposal_${CLAUDE_SESSION_ID}.xlsx"
ANALYSIS_FILE="$DEAL_DIR/analysis_${CLAUDE_SESSION_ID}.pdf"
DECK_FILE="$DEAL_DIR/deck_${CLAUDE_SESSION_ID}.pptx"

Cross-Referencing Sessions

Link extraction sessions with triage and proposal analyses:

# Create session manifest for cross-referencing
SESSION_MANIFEST="deals/${COMPANY_NAME}/session_manifest.json"

# Append session record
cat >> "$SESSION_MANIFEST" << EOF
{
  "session_id": "${CLAUDE_SESSION_ID}",
  "timestamp": "$(date -u +%Y-%m-%dT%H:%M:%SZ)",
  "workflow_type": "full_analysis",
  "related_sessions": ["<previous_triage_session>", "<cim_extraction_session>"],
  "outputs": ["triage.json", "extraction.json", "proposal.xlsx", "analysis.pdf"]
}
EOF

Audit Trail for Due Diligence

Maintain complete audit trails for investment committee review:

# Create audit log with all session activities
AUDIT_LOG="deals/${COMPANY_NAME}/audit_trail.log"

# Log each action with session context
echo "[$(date -u +%Y-%m-%dT%H:%M:%SZ)] SESSION=${CLAUDE_SESSION_ID} ACTION=triage SCORE=8.2 RECOMMENDATION=PROCEED" >> "$AUDIT_LOG"
echo "[$(date -u +%Y-%m-%dT%H:%M:%SZ)] SESSION=${CLAUDE_SESSION_ID} ACTION=extract SOURCE=CIM.pdf PAGES=45" >> "$AUDIT_LOG"
echo "[$(date -u +%Y-%m-%dT%H:%M:%SZ)] SESSION=${CLAUDE_SESSION_ID} ACTION=proposal IRR=35.2 MOIC=4.2" >> "$AUDIT_LOG"

Benefits

• •Complete audit trail: Every analysis tracked per session for compliance
• •Multi-session workflows: Continue analysis across multiple Claude sessions
• •Team handoff: Share session IDs for seamless collaboration
• •Parallel evaluation: Run multiple deal analyses simultaneously
• •Historical reference: Trace back any decision to its source session

Complete Deal Workflow

Phase 1: Initial Screening

# 1. Set up deal tracking
DEAL_DIR="deals/CompanyName_${CLAUDE_SESSION_ID}"
mkdir -p "$DEAL_DIR"

# 2. Triage the deal
/mna triage CompanyName revenue=50 ebitda=15 growth=25 > "$DEAL_DIR/triage.json"

# If score >= 7, proceed

Phase 2: Deep Analysis

# 2. Extract CIM data
/mna extract CompanyName_CIM.pdf

# 3. Optional: AI cost analysis
/mna aimpact CompanyName_financials.xlsx

Phase 3: Financial Modeling

# 4. Generate proposal (need IRR >= 20%)
/mna proposal CompanyName

# 5. Create investment analysis PDF
/mna analysis CompanyName

Phase 4: Board Approval

# 6. Generate board deck
/mna deck CompanyName --format full

Implementation Notes

MCP Tools (if available)

• •mcp__nuvini-mna__triage_deal
• •mcp__nuvini-mna__generate_proposal
• •mcp__nuvini-mna__create_presentation

Script Locations

• •Triage: /Volumes/AI/Code/MNA/nuvini-ma-system-complete/triage-analyzer/
• •Proposal: /Volumes/AI/Code/MNA/nuvini-ma-system-complete/mna-proposal-generator/
• •Presenter: /Volumes/AI/Code/MNA/nuvini-ma-system-complete/committee-approval-presenter/
• •Extractor: skills/mna-toolkit/scripts/

API Endpoints (if running)

• •POST /api/triages - Create triage
• •POST /api/triages/{id}/generate-investment-analysis - Generate PDF
• •GET /download/investment-reports/{filename} - Download report

Related Skills

This toolkit consolidates:

• •triage-analyzer → /mna triage
• •financial-data-extractor → /mna extract
• •mna-proposal-generator → /mna proposal
• •investment-analysis-generator → /mna analysis
• •committee-presenter → /mna deck
• •aimpact → /mna aimpact

The individual skills remain available for backward compatibility.