npbuilds Diligence Scorecard — The Integration Engine
This is the capstone skill of the Asclepius domain. Every other skill produces a component analysis; the diligence scorecard synthesizes all of them into the artifact that venture investors actually produce — a structured, scored assessment with an explicit recommendation. No open-source biotech diligence scorecard exists.
The 8-pillar framework is deliberately opinionated. Seven pillars are standard in biotech diligence. The 8th — Computational Infrastructure — reflects the $1B+ mega-rounds flowing to AI science factories (Xaira $1B, Isomorphic $600M, Lila Sciences $350M) and the growing requirement that founding teams have computational expertise (Dimension Capital). This is where your Formation Bio experience creates maximum differentiation.
How to Run
Input
The scorecard consumes outputs from all other Asclepius skills:
| Input Source | Pillar(s) Fed |
|---|---|
| pos-calculator | Pillar 1 (Clinical Strength) |
| endpoint-selection | Pillar 1, Pillar 2 |
| trial-design-optimizer | Pillar 1 |
| regulatory-precedent + pathway-analyzer | Pillar 2 (Regulatory Positioning) |
| regulatory-risk-scorer | Pillar 2 |
| clinical-differentiator + pipeline-mapper | Pillar 3 (Competitive Position) |
| market-dynamics | Pillar 3 |
| cmc-risk-assessor + modality-manufacturing | Pillar 4 (Manufacturing Feasibility) |
| patent-analyzer + ip-valuation | Pillar 5 (IP Fortress) |
| rnpv-modeler + peak-sales-forecaster + deal-economics | Pillar 6 (Financial Attractiveness) |
| User assessment / public information | Pillar 7 (Team & Execution) |
| User assessment / public information | Pillar 8 (Computational Infrastructure) |
If upstream analyses haven't been run, the scorecard can operate with partial information — but must flag which pillars are scored on limited data.
Steps
Step 1 — Score Each Pillar (1-10)
Use the scoring calibration in references/scoring-calibration.md for criteria at each level.
Pillar 1: Clinical Strength (Weight: 20%) Assess: PoS from pos-calculator, endpoint quality from endpoint-selection, data maturity, biomarker strategy. Key question: "Is the clinical evidence compelling enough to justify Phase 3 investment?"
Pillar 2: Regulatory Positioning (Weight: 15%) Assess: Pathway from pathway-analyzer, regulatory risk from regulatory-risk-scorer, designation status, precedent depth. Key question: "Does the regulatory path to approval have precedent and manageable risk?"
Pillar 3: Competitive Position (Weight: 15%) Assess: Differentiation from clinical-differentiator, landscape from pipeline-mapper, market dynamics. Key question: "Can this asset win meaningful market share against current and future competitors?"
Pillar 4: Manufacturing Feasibility (Weight: 10%) Assess: CMC risk from cmc-risk-assessor, COGS from modality-manufacturing, scalability. Key question: "Can this drug be manufactured reliably at commercial scale at acceptable COGS?"
Pillar 5: IP Fortress (Weight: 10%) Assess: Patent landscape from patent-analyzer, IP contribution from ip-valuation, exclusivity periods. Key question: "Is the IP strong enough to protect commercial returns through peak sales?"
Pillar 6: Financial Attractiveness (Weight: 15%) Assess: rNPV from rnpv-modeler, peak sales from peak-sales-forecaster, deal terms from deal-economics. Key question: "Does the risk-adjusted return justify the capital at risk?"
Pillar 7: Team & Execution (Weight: 10%) Assess: Management experience, board composition, operational capability, capital adequacy. Key question: "Can this team execute the development plan and navigate setbacks?"
Pillar 8: Computational Infrastructure (Weight: 5%) Assess: AI/ML platform maturity, proprietary data assets, wet-lab/dry-lab integration, computational team depth. Key question: "Does this company have computational capabilities that accelerate development or create data moats?"
Step 2 — Check for Red Flags
Red flags automatically cap the aggregate score at 4.0:
- Going concern risk (<6 months cash, no financing path)
- Data integrity concerns
- Regulatory rejection (CRL) without clear resubmission path
- Core patent invalidated
- Management exodus (CEO + CSO departed)
Step 3 — Check for Green Flags
Green flags add +0.5 to aggregate (cap at 10.0):
- Significant insider buying (>$500K open market)
- Top-tier VC validation (Arch, Flagship, Third Rock)
- Big pharma co-development partnership
- Platform with multiple advancing indications
Step 4 — Calculate Aggregate Score
Aggregate = Σ (Pillar Score × Weight) + Green Flag Bonus - Red Flag Cap
Step 5 — Generate Recommendation
| Aggregate | Recommendation | Position Sizing |
|---|---|---|
| 8.0-10.0 | Strong Buy | Maximum position (3-5% of fund) |
| 6.5-7.9 | Buy | Standard position (1-3% of fund) |
| 5.0-6.4 | Hold | Monitor for catalysts; small position if specific thesis |
| 3.5-4.9 | Cautious | Below average; avoid unless compelling contrarian thesis |
| 1.0-3.4 | Avoid | Fundamental concerns; do not invest |
Output
DILIGENCE SCORECARD — [Company / Asset Name]
Indication: [therapeutic area + indication]
Current Phase: [development stage]
Date: [assessment date]
Analyst: [name]
═══════════════════════════════════════════════
PILLAR SCORES
Score Weight Weighted
1. Clinical Strength [X/10] 20% [X.X]
2. Regulatory Positioning [X/10] 15% [X.X]
3. Competitive Position [X/10] 15% [X.X]
4. Manufacturing Feasibility [X/10] 10% [X.X]
5. IP Fortress [X/10] 10% [X.X]
6. Financial Attractiveness [X/10] 15% [X.X]
7. Team & Execution [X/10] 10% [X.X]
8. Computational Infrastructure [X/10] 5% [X.X]
─────────────────────
AGGREGATE: [X.X / 10.0]
Red Flags: [None / List]
Green Flags: [None / List]
Adjusted Aggregate: [X.X / 10.0]
═══════════════════════════════════════════════
RECOMMENDATION: [Strong Buy / Buy / Hold / Cautious / Avoid]
One-Line Thesis: [Single sentence investment thesis]
Key Catalysts:
1. [Catalyst + expected date + value impact]
2. [Catalyst + expected date + value impact]
3. [Catalyst + expected date + value impact]
Key Risks:
1. [Risk + probability + impact + mitigant]
2. [Risk + probability + impact + mitigant]
3. [Risk + probability + impact + mitigant]
Financial Summary:
rNPV: $[X]M (range: $[low] - $[high])
Peak Sales: $[X]B
PoS (LOA): [X]%
Next Value Inflection: [event] ([X]x potential)
Pillar-Level Detail:
Strongest Pillar: [name] ([score]) — [why]
Weakest Pillar: [name] ([score]) — [why, and what would improve it]
═══════════════════════════════════════════════
DATA COMPLETENESS
Pillars with full upstream analysis: [list]
Pillars scored on limited data: [list + what's missing]
Confidence in aggregate: [High / Medium / Low]
Error Handling
| Scenario | Response |
|---|---|
| Missing upstream analysis for a pillar | Score based on available information; flag as "limited data" in output; widen confidence range |
| Private company (limited public data) | Pillars 7-8 may rely on investor deck + news; note data source limitations |
| Pre-clinical company | Pillars 1, 2, 3 scored conservatively; emphasize platform/team over clinical data |
| Multi-asset company | Score lead asset as primary; note pipeline optionality as green flag |
| User disagrees with a pillar score | Pillar weights are configurable; user can adjust weight to 0% to exclude a pillar |
Cross-Domain Connections
- All biotech-venture skills: This is the integration point — every other skill feeds into at least one pillar
- Biotech-venture/investment-memo-writer: Scorecard outputs feed directly into IC memo generation
- Biotech-venture/portfolio-analyzer: Individual scorecards feed portfolio-level analysis
- Investing/archon: Scorecard parallels Archon's investment thesis evaluation framework
- Investing/second-level-thinking: Red/green flag analysis applies contrarian lens