Stats
Actions
Tags
From deal-output
Write a structured, publication-ready Investment Committee memo for a PE buyout, acquisition, or strategic investment. Use this skill whenever Ian asks to "write the IC memo," "draft the investment memo," "build the IC package," "write up the deal," "draft the investment thesis memo," or "put together the IC write-up." Also triggers for: "memo for IC," "investment committee write-up," "deal memo," "write up the thesis," "IC-ready memo." This skill governs IC memo document architecture — section structure, length calibration, how the Six Screening Questions map to memo sections, and formatting via pattern-docx. It does not govern: analytical methodology (mckinsey-consultant), Six Screening Questions content (investment-evaluation-framework.md), claim standards (writing-style), logic redlining (claim-scrutinizer), adversarial stress-testing (red-team), or moat evidence methodology (competitive-moat-assessment). Defer to those skills for their respective responsibilities.
How this skill is triggered — by the user, by Claude, or both
Slash command
/deal-output:ic-memoThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
You are writing an Investment Committee memo. Read this entire file before beginning.
You are writing an Investment Committee memo. Read this entire file before beginning.
An IC memo is not a market research report. It is not a strategy deck. It is a document that makes a specific, evidence-backed recommendation to a small group of experienced investors who will challenge every assertion. The governing discipline is precision under pressure — every claim must survive a hostile question from an IC member who has read the same CIM and talked to the same management team.
ic-memo ← YOU ARE HERE — document architecture, section structure, formatting
│
├── investment-evaluation-framework.md ← Six Screening Questions as gate structure
│ Read: /mnt/skills/user/mckinsey-consultant/references/investment-evaluation-framework.md
│
├── mckinsey-consultant ← analytical methodology, pyramid principle, SCR
│
├── ntb-diligence ← (OPTIONAL UPSTREAM) standalone NTB diligence
│ Run FIRST when the deal warrants structured NTB analysis.
│ Produces NTB registry + diligence plan + stress test that feed Sections 4, 7, 9.
│ When ntb-diligence has run, do NOT re-derive NTBs inline — consume its output.
│ Read: /mnt/skills/user/ntb-diligence/SKILL.md
│
├── market-research ← evidence gathering for Sections 4, 5, 6
│ Runs in IC memo mode: no standalone intake, no standalone output
│ Pyramid levels: L4 (market) → L3 (customer) → L2 (competitive) only
│ L1 (company position) handled by ic-memo directly via intake materials
│ Theme count: 3–4 (not 4–6) — IC memo is tighter than a standalone report
│ Output: evidence inputs to memo sections, not a standalone deliverable
│
├── competitive-moat-assessment ← moat evidence methodology (load after L2 research)
│ Uses L2 findings as inputs — run after market-research L2 completes
│
├── writing-style ← prose standards, inductive chain, claim tagging
│ Run self-review before generating any file
│
├── claim-scrutinizer ← seven-part test, assumption audit, base rates
│ Run full redline after first draft
│
├── red-team ← adversarial pass using investment attack lenses
│ Run after claim-scrutinizer hardening
│
├── pre-mortem ← failure mode inventory with NTB mapping
│ Run after red-team; extends stress test coverage
│
├── pattern-docx ← file output
│ doc-quality-checker ← run immediately after file delivery
└── kpi-tree-builder ← post-close: converts IC thesis + NTB registry into trackable operating architecture
Load order: Read investment-evaluation-framework.md first. Then follow this skill for document architecture. Defer to named skills for their responsibilities.
market-research in IC memo context: market-research runs in a constrained mode within the IC memo workflow. It does not run its own intake protocol (ic-memo already captured the required inputs). It does not produce a standalone deliverable. It produces evidence — sourced, cited, triangulated findings — that feed directly into memo sections. This is the critical distinction: market-research is the evidence layer; ic-memo is the document layer.
ntb-diligence in IC memo context: ntb-diligence is optional but recommended for any deal where the investment thesis rests on 4+ distinct assumptions with meaningful MOIC impact each. When it has been run upstream of this skill, its output is the authoritative source for the NTB framing — the NTB registry feeds Section 4 (thesis pillars are derived from NTBs), the diligence plan feeds Section 10 open items, and the stress tests feed Section 9 (risks) with ready-made leading indicators and kill criteria. Do NOT re-derive the NTBs inline when ntb-diligence has already produced them.
Load the investment evaluation framework:
Read: /mnt/skills/user/mckinsey-consultant/references/investment-evaluation-framework.md
Check for a matching domain template:
Scan the company or topic against /mnt/skills/user/ic-memo/references/domain-templates/
for any file whose trigger list matches. If a match is found, load it immediately — it
pre-loads confirmed data, prior NTB diligence output, Gate 2 scorecards, pre-mortem
failure modes, and open questions that would otherwise require re-gathering.
Currently available IC memo domain templates:
sea-ltd-sea-brazil.md — Sea Limited (NYSE: SE), Shopee, Monee, SPX Express, Garena,
SEA e-commerce investment, MELI vs Sea, TikTok Shop SEA, any SEA country e-commerceState in the intake output: "Domain template loaded: [filename]. [N] confirmed data points pre-loaded. [N] open questions carried forward from prior analysis."
Check for ntb-diligence upstream output: If ntb-diligence has already been run on this deal (either in the current session or referenced as prior work), treat its Phase 3 NTB registry, Phase 3 diligence plan, and Phase 4 stress tests as authoritative inputs. Do NOT re-derive the NTBs during drafting. The NTB registry becomes the structural frame for Section 4 thesis pillars; the diligence plan feeds Section 10 open items; the stress tests feed Section 9 risks.
Then collect the following. Do not begin writing until all critical inputs are in hand.
Required:
Helpful but not blocking:
If a CIM or research document is attached, read it fully before asking any questions. Extract what you can; ask only for genuine gaps.
When this skill is triggered without sufficient context, ask all of the following in one structured message. Do not ask questions one at a time.
Before I draft the IC memo, I need a few inputs:
1. The company Name, what they do, who they sell to, revenue/ARR scale.
2. The deal Type (buyout / acquisition / minority), entry valuation, deal structure, hold period.
3. The thesis Your working investment thesis — what makes this a good investment at this price? Even a rough version is fine; I'll pressure-test and sharpen it.
4. The decision context Is this a first look, a go/no-go before LOI, or a final IC presentation? This determines depth and emphasis.
5. Known risks or IC concerns What are the 1–2 things that will face the hardest questions at IC?
6. Existing materials Attach or paste any CIM, management deck, research, or prior memos.
After receiving inputs:
A Pattern IC memo has nine sections in this exact order. Do not reorder or combine. Section length calibrations assume a final memo of 8–12 pages for a standard buyout. Adjust proportionally for first-look (4–6 pages) vs. final IC (12–15 pages).
1. Cover Block
2. Executive Summary (1 page)
3. Company Overview (0.5–1 page)
4. Investment Thesis (1–1.5 pages)
5. Market & Competitive (1–1.5 pages)
6. Business Quality (1.5–2 pages)
7. Financial Analysis (1–1.5 pages)
8. Deal Structure & Returns (1 page)
9. Risks & Mitigants (1 page)
10. IC Recommendation (0.5 page)
Before writing any memo sections, execute the market-research pyramid for Levels 4, 3, and 2. This produces the evidence base that Sections 4, 5, and 6 draw from. Without this phase, those sections will rely on CIM-sourced claims that have not been independently triangulated — which will fail claim-scrutinizer Pass 2.
Read: /mnt/skills/user/market-research/SKILL.md → IC Memo Mode (below)
market-research runs in a constrained mode here. Three differences from standalone:
1. No intake protocol The ic-memo intake already captured: company, market, deal context, key questions, and known data challenges. Do not run market-research's own intake. Pass the ic-memo intake responses directly as context.
2. Scoped pyramid — L4, L3, L2 only
3. No standalone deliverable market-research does not produce a report or deck here. It produces a structured evidence summary — findings organized by pyramid level with inline citations ([Source, Year] [H/M/L]) — that feeds directly into memo drafting. Theme development is condensed: 3–4 themes maximum, structured around the investment thesis pillars rather than independent structural observations.
After market-research completes, organize findings as:
MARKET RESEARCH EVIDENCE SUMMARY — [Company Name]
L4 FINDINGS (feeds Sections 4 + 5 market block):
- Market size: [figure] [source] [confidence]
- CAGR: [figure] [source, base year] [confidence]
- Trend 1: [finding] → [implication for thesis]
- Trend 2: [finding] → [implication for thesis]
- DATA GAPS: [any flagged gaps]
L3 FINDINGS (feeds Section 6b):
- Segment 1: [JTBD framing] [evidence]
- Segment 2: [JTBD framing] [evidence]
- Primary barrier: [finding] [evidence]
- DATA GAPS: [any flagged gaps]
L2 FINDINGS (feeds Section 5 competitive block + competitive-moat-assessment):
- Competitor 1: [6-element profile]
- Competitor 2: [6-element profile]
- Positioning: [white space finding]
- DATA GAPS: [any flagged gaps]
Once this summary is produced, run competitive-moat-assessment using the L2 findings as inputs before beginning memo drafting.
The 10-section structure below is the default prescribed architecture for a PE buyout or acquisition memo. Public equity long memos and certain deal structures may use a compressed 8-section variant that combines some sections:
| Prescribed (PE buyout) | Compressed (public equity long) |
|---|---|
| 1. Cover Block | 1. Cover Block |
| 2. Executive Summary | 2. Executive Summary |
| 3. Company Overview | 3. Company Overview |
| 4. Investment Thesis | 4. Investment Thesis (with NTB registry) |
| 5. Market & Competitive Position | 5. Market & Competitive Position |
| 6. Business Quality | (merged into Section 4 Investment Thesis) |
| 7. Financial Analysis | 6. Unit Economics & Valuation (combined) |
| 8. Deal Structure & Exit | (N/A for public equity long) |
| 9. Risks & Mitigants | 7. Risk Analysis & Pre-Mortem |
| 10. IC Recommendation | 8. Investment Recommendation |
When to use compressed 8-section: Public equity long positions where Deal Structure & Exit is not a discrete negotiation item (you buy shares at market), and where Business Quality is most naturally integrated with the Investment Thesis pillars rather than stood up as a separate section.
When to use prescribed 10-section: PE buyout, acquisition, or strategic investment where deal structure is a live negotiation, where Gates 4-5 (Exit Realization and Owner Fit) require their own treatment, and where Gate 1 (Company Quality) warrants a full Business Quality section distinct from the thesis argument.
The discipline is the same for both. The Six Screening Questions map to sections in each architecture; claim standards, reconciliation, and output passes apply identically. The divergence is structural, not analytical.
If you are unsure which architecture applies, ask before drafting — the answer depends on deal type and Ian's preference for the specific investment.
Pattern-docx Normal paragraphs with run-level formatting. No heading style.
[Company Name] 36pt, SemiBold, color 1F4E79
Investment Committee Memorandum 20pt, color 2E75B6
[Deal type] | [Date] | CONFIDENTIAL italic, color 595959
[blank line]
CONFIDENTIAL — FOR AUTHORIZED RECIPIENTS ONLY 9pt, Bold, color C00000
This is the most important section. IC members read this first and many read only this. If the executive summary is unclear or buried, the memo fails regardless of the quality of subsequent sections.
Invoke executive-summary-writer skill for this section:
Read: /mnt/skills/user/executive-summary-writer/SKILL.md → Format A (One-Page Memo)
Structure (five blocks in order):
The executive summary must be self-contained. A reader who reads only this section must be able to answer: what is the conclusion, why is it correct, what is the biggest risk, and what happens next.
Factual orientation. Not analytical. Sets the stage for the thesis.
Required elements:
Length discipline: This section has a hard ceiling of 1 page. If it runs longer, content belongs in the Business Quality section or an appendix.
Claim standard: All figures cited here are facts — sourced from CIM, management presentation, or public filings. No estimates in this section without explicit labeling.
The affirmative case for the investment, structured as a pyramid.
Structure:
Pillar construction standard: Each pillar follows inductive chain discipline (governed by writing-style):
A pillar that skips any link in the chain is an assertion, not an argument.
Label every claim: [F] fact / [E] estimate / [H] hypothesis
Gate coverage: The three pillars must collectively address at minimum:
Gates 4–6 are addressed in subsequent sections. Do not try to cover all six gates in the thesis section — that produces an unfocused document.
Evidence source: Draw from the Market Research Evidence Summary produced in the Market Research Phase above. Do not re-research here — compress and synthesize the L4 and L2 findings into the memo's length constraints.
Structure:
For competitive position — invoke competitive-moat-assessment using L2 findings:
Read: /mnt/skills/user/competitive-moat-assessment/SKILL.md
competitive-moat-assessment receives the L2 competitor profiles as inputs and produces the moat verdict. This verdict — not the competitor profiles themselves — is the primary output of the competitive section.
This section must name competitors and state their positions. Generic competitive landscape descriptions ("the market is fragmented with many players") do not pass.
Gate coverage: Addresses Gate 2 (Sector Timing) and Gate 1 competitive position sub-question.
The deepest analytical section. Addresses Gate 1 in full.
Required subsections:
6a. Business Model & Unit Economics
6b. Customer Quality
6c. Management Assessment
6d. Competitive Moat (summary — full analysis in Section 5)
Gate coverage: Addresses Gate 1 (Company Quality) in full.
Required elements:
7a. Historical Performance
7b. Base Case Model
7c. Return Analysis
Gate coverage: Addresses Gate 3 (Investment Attractiveness) in full.
Structure:
8a. Deal Structure
8b. Exit Analysis
Gate coverage: Addresses Gate 4 (Exit Realization) and Gate 5 (Owner Fit).
Format: Risk register table — not a narrative list.
| # | Risk | P | M | Score | Mitigant | Adequacy |
|---|---|---|---|---|---|---|
| 1 | [specific risk] | 1–5 | 1–5 | P×M | [specific mitigant] | Adequate/Partial/Insufficient |
P = Probability (1=unlikely, 5=very likely) M = Magnitude (1=negligible, 5=thesis-killing) Score = P × M (range 1–25)
Risk selection rules:
Gate coverage: Addresses Gate 6 (Adversarial Diligence).
Four mandatory elements — all required, no exceptions:
1. Recommendation statement Direct, binary: "We recommend proceeding to LOI" / "We recommend passing at this price" / "We recommend conditional approval subject to [specific conditions]." No hedging. No "we believe this merits further consideration."
2. Walk-away conditions Specific binary findings that, if confirmed, cause the recommendation to reverse. Stated as observable facts, not risk categories.
3. Key assumption The single assumption the recommendation rests on, labeled explicitly as an assumption. Named — not "market growth" but the specific figure, mechanism, and the evidence that currently supports it.
4. Open items before proceeding The specific questions that must be answered before a yes-vote is defensible. Named questions, named owners, named deadlines — not "conduct further diligence."
Gate coverage: Synthesizes all six gates into a final binary verdict.
Run all five passes before delivering the final file.
Pass 1 — writing-style self-review Draft all prose in plain text first. Run writing-style Steps 1–5 on every section. Fix all flags before generating any code. Focus especially on: inductive chains in Section 4 thesis pillars, exclusivity terms in moat claims, timeline claims in financial analysis. Pass 1 also catches Group E draft artifact language — no "v[N]" labels on cover, no "(NEW)" tags in section headers, no "pre-mortem addition:" prefixes in body text, no FM codes in body text.
Read: /mnt/skills/user/writing-style/SKILL.md
Pass 2 — claim-scrutinizer redline Full seven-part test on every material claim. Mandatory focus areas:
Read: /mnt/skills/user/claim-scrutinizer/SKILL.md
Pass 3 — red-team adversarial pass Load Type A investment attack lenses. Attack load-bearing pillars first.
Read: /mnt/skills/user/red-team/SKILL.md
Read: /mnt/skills/user/red-team/references/red-team-investment-attacks.md
Pass 4 — pre-mortem (failure mode inventory) Run after red-team hardening. Pre-mortem assumes the deal has already failed and works backward through failure modes, extending the risk register with named mechanisms, leading indicators, boundability assessments, and compound failure paths.
Read: /mnt/skills/user/pre-mortem/SKILL.md
Pass 4b — Numeric Consistency Reconciliation (mandatory after pre-mortem)
The pre-mortem produces dollar and MOIC impact numbers in every material failure mode deep dive. These numbers must reconcile to the valuation section — otherwise the document contradicts itself internally.
Produce the reconciliation before Pass 5:
From Section 7 (Financial Analysis) or wherever the valuation lives, extract the authoritative Base Assumptions Table: entry equity, exit multiple, hold period, base case FY exit EBITDA and GMV, contribution margin rules.
For every pre-mortem 9-field deep dive's Failure Spectrum field, verify:
For every numeric claim that appears in more than one section (e.g., "$159B FY2029E GMV" in both investment thesis and valuation), verify the value is identical everywhere it appears. Flag any divergence.
For every multi-section metric with distinct period (e.g., "$127B FY2025 GMV" vs. "$159B FY2029E GMV"), verify the period label is explicit in every occurrence.
If any check fails, fix before Pass 5. Cross-section numeric contradictions are more damaging than draft artifact language and must be resolved first.
This pass is owned by the pre-mortem skill's Section 6 ↔ Section 7 Reconciliation procedure but must be explicitly re-run at the IC memo level because the pre-mortem author may have drafted impact fields before the valuation section was finalized.
Pass 4c — boundability (convert risk to underwriting action)
Run after Pass 4 (pre-mortem) and Pass 4b (numeric reconciliation). Boundability takes the failure mode registry produced by pre-mortem and converts each material diligence item into specific underwriting actions across five buckets (model, price, leverage, docs, operations) plus a deal-level verdict.
The unit of assessment in boundability is typically the NTB, not the individual failure mode — this keeps the output organized around the thesis claims the IC will debate rather than the underlying risk mechanisms. Each NTB receives one issue object with six module scores (Perimeter, Timing, Data Quality, Outcome Range, Precedent/Observability, Mitigants) and a final classification (Boundable / Partially Boundable / Unboundable).
Boundability terminology matches pre-mortem (deliberately). An NTB classified Boundable here must score ≥25/30 on the six modules; an NTB classified Partially Boundable or Unboundable by pre-mortem should score <25. If pre-mortem and boundability disagree on classification for the same item, resolve before delivery.
For public equity long positions (where there is no deal negotiation), the Leverage and Docs buckets collapse; Price becomes "entry discipline" and Operations becomes "monitoring cadence." For PE control buyouts, all five buckets are fully active.
Read: /mnt/skills/user/boundability/SKILL.md
Pass 5 — doc-quality-checker Run immediately after file delivery. Do not wait to be asked. Catches any remaining Group E draft artifact language that survived Pass 1 (version labels, changelog subtitles, "pre-mortem addition:" prefixes, FM codes in body text) — all rated at 🔴 CRITICAL.
Read: /mnt/skills/user/doc-quality-checker/SKILL.md
Before generating any code:
Read: /mnt/skills/user/pattern-docx/SKILL.md
DDR-specific formatting notes:
The memo is ready for IC only when all items below pass.
Analytical integrity
Document structure
Iteration loop
Read: /mnt/skills/user/mckinsey-consultant/references/investment-evaluation-framework.md
Read: /mnt/skills/user/ntb-diligence/SKILL.md (optional upstream)
Read: /mnt/skills/user/competitive-moat-assessment/SKILL.md
Read: /mnt/skills/user/executive-summary-writer/SKILL.md
Read: /mnt/skills/user/writing-style/SKILL.md
Read: /mnt/skills/user/claim-scrutinizer/SKILL.md
Read: /mnt/skills/user/red-team/SKILL.md
Read: /mnt/skills/user/red-team/references/red-team-investment-attacks.md
Read: /mnt/skills/user/pre-mortem/SKILL.md
Read: /mnt/skills/user/pattern-docx/SKILL.md
Read: /mnt/skills/user/doc-quality-checker/SKILL.md
Domain templates (auto-load on matching trigger):
/mnt/skills/user/ic-memo/references/domain-templates/sea-ltd-sea-brazil.md
npx claudepluginhub ian-lawrence423/claude-skills --plugin deal-outputProvides a checklist for code reviews covering functionality, security, performance, maintainability, tests, and quality. Use for pull requests, audits, team standards, and developer training.