deal-screener
Screen PE and CRE deal opportunities. Apply scoring criteria, filter by investment thesis, rank targets by fit. Use when evaluating new deals, screening pipeline opportunities, or prioritizing investm
SKILL.mdskills/deal-screener/SKILL.md
GitHub/deal-screener -- Deal Opportunity Screener
Apply scoring criteria to PE/CRE targets and output ranked recommendations.
Before executing, read
_shared/plugin-context.mdfor Nexus domain vocabulary, MCP tool references, and quality standards.
Prerequisites
Required: None -- this skill works with manually provided deal data.
Optional MCP Connections (enhance automation):
- Nexus MCP (
nexus.syntari-ai.com/api/mcp) --nexus_list_deals,nexus_get_deal,nexus_screen_dealfor automated deal retrieval and scoring - BigData.com MCP (via
financial-analysisplugin) -- company tearsheets, financials, sentiment for PE target research - PitchBook MCP -- private market deal data, comparable transactions
- S&P Global / Kensho MCP -- Capital IQ financials, company profiles
Graceful Degradation: If MCP tools are not connected, provide deal data manually (paste deal summaries, upload spreadsheets, or describe opportunities in conversation). All analytical frameworks work identically with manual input.
When to Use
- Evaluating a batch of new deal opportunities
- Filtering the deal pipeline by investment criteria
- Ranking targets for investment committee review
- Running weekly pipeline scoring refresh
- Comparing inbound proprietary deals against market benchmarks
- Generating pass/proceed recommendations with evidence
Scoring Framework
PE Deal Scoring (100 points)
| Factor | Weight | Criteria |
|---|---|---|
| Revenue quality | 20 | Recurring %, growth rate, concentration |
| Market position | 15 | TAM, market share, competitive moat |
| Management team | 15 | Track record, retention, depth |
| Financial health | 20 | Margins, cash flow, debt levels |
| Growth potential | 15 | Organic + M&A opportunities |
| Valuation | 15 | Multiple vs comps, entry price |
CRE Deal Scoring (100 points)
| Factor | Weight | Criteria |
|---|---|---|
| Location quality | 20 | Submarket fundamentals, demographics |
| Asset quality | 15 | Condition, vintage, deferred maintenance |
| Income stability | 20 | Occupancy, WALT, tenant credit |
| Return profile | 20 | Cap rate, IRR, cash yield |
| Value-add potential | 15 | Rent upside, repositioning opportunity |
| Risk factors | 10 | Environmental, regulatory, market cycle |
Signal Output
| Score | Signal | Color |
|---|---|---|
| 80-100 | Strong Buy | Green (#10B981) |
| 60-79 | Buy | Blue (#068EFF) |
| 40-59 | Hold | Amber (#F59E0B) |
| 0-39 | Pass | Red (#EF4444) |
Data Shapes
Deal Screen Input
interface DealScreenInput {
dealId?: string // Optional -- omit if providing data manually
dealName: string
dealType: 'pe' | 'cre'
criteria?: {
minRevenue?: number // Minimum revenue threshold ($)
maxMultiple?: number // Maximum entry multiple (EV/EBITDA)
sectors?: string[] // Target sectors (e.g., ["SaaS", "Health IT"])
geographies?: string[] // Target geographies (e.g., ["US", "Western Europe"])
minScore?: number // Minimum score threshold for advancement
}
}
Deal Screen Result
interface DealScreenResult {
dealId?: string
dealName: string
dealType: 'pe' | 'cre'
totalScore: number // 0-100 weighted score
signal: 'strong_buy' | 'buy' | 'hold' | 'pass'
signalColor: string
factorScores: {
factor: string // e.g., "Revenue quality"
weight: number // e.g., 20
score: number // 0-100 raw score for this factor
weightedScore: number // weight * score / 100
evidence: string // Supporting evidence for the score
}[]
recommendation: string
keyHighlights: string[]
keyRisks: string[]
}
Workflow
Step 1: Gather Deal Data
- With Nexus MCP: Use
nexus_list_dealsornexus_get_dealto retrieve deal data automatically. - With BigData.com MCP: Use
bigdata_company_tearsheetorbigdata_searchfor company research. - Without MCP: Ask the user to provide deal details (company name, sector, financials, property info).
Step 2: Define Criteria
Confirm investment thesis parameters (sector, size, geography, return threshold).
Step 3: Score Each Deal
Apply the scoring framework to each opportunity. Document evidence for each factor score.
Step 4: Rank and Filter
Sort by total score. Apply minimum threshold (typically 60+ for further diligence).
Step 5: Output
Generate ranked table with scores, signal colors, and key highlights per deal.
Sample Output
=== Deal Screening Report -- Q1 2026 Pipeline ===
Fund: SVP Growth Equity Fund IV
Criteria: Software/SaaS, $50M-$500M EV, >15% revenue growth
Rank Deal Sector EV Score Signal Key Factor
------------------------------------------------------------------------------
1 Meridian Data SaaS $420M 87.4 STRONG BUY 92% recurring, 118% NRR
2 Apex Cloud Platform SaaS $285M 82.1 STRONG BUY 40% YoY growth, net neg churn
3 NovaCare Analytics Health IT $195M 74.6 BUY FDA-cleared, 87% gross margin
4 TerraForm Industrial SaaS $340M 68.3 BUY Strong TAM, mgmt concern
5 CloudBridge CRM SaaS $120M 55.2 HOLD Revenue concentration risk
6 DataSync Corp Data $250M 41.8 HOLD Declining margins, stale product
7 Legacy Systems Inc IT Svcs $180M 28.4 PASS No recurring revenue, 5% growth
Summary: 7 deals screened -> 2 Strong Buy, 2 Buy, 2 Hold, 1 Pass
Recommendation: Advance Meridian Data and Apex Cloud to Phase II diligence
Workflow Chain
This skill is step 2 in the deal analysis chain: /deal-doc-triage -> /deal-screener -> /deal-memo-writer -> deep diligence.
Related Skills
/deal-memo-writer-- Write PE acquisition and CRE investment deal memos from screening results/financial-modeling-- LBO, DCF, M&A models to support deal valuation/competitive-analysis-- Analyze competitors of screening targets/om-parser-- Parse CRE offering memorandums for deal data extraction/cre-underwriter-- Full CRE underwriting for screened CRE deals
Confidence Calibration
Every screening output must include a calibrated confidence assessment:
## Confidence Assessment
- **Score**: 0.XX (0.0 = no confidence, 1.0 = full confidence)
- **Calibration**: conservative | neutral | optimistic
- **Confidence Drivers**: [what supports the score -- e.g., "complete financials provided", "multiple comp transactions available"]
- **Data Gaps**: [what's missing that would improve confidence -- e.g., "no management references", "market rent data unavailable"]
Three-pass calibration: Generate scoring from conservative, neutral, and optimistic perspectives. High consistency across passes = high confidence. Divergence = flag for human review.
Typical confidence ranges:
- 0.85+: Complete financial data, multiple comps, management access
- 0.65-0.84: Partial data, limited comps, preliminary assessment
- 0.40-0.64: Teaser/headline data only, significant assumptions required
- Below 0.40: Insufficient data -- flag for additional information gathering before scoring