Institutional Equity Analyst

View saved version →

---
name: accounting-analysis
description: Use for three-statement accounting, revenue recognition, margins, working capital, capex, stock compensation, leases, goodwill, taxes, pensions, and accounting normalization.
---

# Accounting Analysis

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M011, M012, M013, M014, M015, M016, M017, M018, M019, M020
- Appendices: Appendix B, Appendix C
- Master Labs: Master Lab 01, Master Lab 02, Master Lab 14
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Domain workflow

1. Read `references/operating-rules.md` and every bundled module reference that is material to the user's request.
2. Define the narrow decision question and the variables capable of changing value or risk.
3. Gather current primary evidence first when the task concerns a live public company.
4. Execute the explicit operating procedures in the loaded modules.
5. Preserve calculations, definitions, assumptions, contrary evidence, and source provenance.
6. Apply each loaded module's failure tests and completion test before calling the work complete.
7. Return a concise answer for a narrow question, but retain a reproducible evidence trail for every material claim.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 19

View saved version →

---
name: alternative-data
description: Use for expert calls, customer or supplier diligence, web and app data, hiring signals, pricing and inventory scraping, channel checks, and validation of alternative data.
---

# Alternative Data

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M056, M057, M058, M059, M060
- Appendices: Appendix F
- Master Labs: Master Lab 07, Master Lab 08
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Domain workflow

1. Read `references/operating-rules.md` and every bundled module reference that is material to the user's request.
2. Define the narrow decision question and the variables capable of changing value or risk.
3. Gather current primary evidence first when the task concerns a live public company.
4. Execute the explicit operating procedures in the loaded modules.
5. Preserve calculations, definitions, assumptions, contrary evidence, and source provenance.
6. Apply each loaded module's failure tests and completion test before calling the work complete.
7. Return a concise answer for a narrow question, but retain a reproducible evidence trail for every material claim.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 12

View saved version →

---
name: analyst-training
description: Use for zero-to-expert analyst training, deliberate practice, certification, drills, casebooks, reviewer question banks, and mastery assessment.
---

# Analyst Training

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M069, M070
- Appendices: Appendix A, Appendix J, Appendix K, Appendix M
- Master Labs: Master Lab 12, Master Lab 13, Master Lab 14
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Training workflow

1. Diagnose the learner's current level from demonstrated work, not tenure or self-rating.
2. Assign progressive filing, accounting, modeling, valuation, industry, forensic, risk, and communication exercises from the bundled references.
3. Require reproducible work products, closed-book drills where specified, reviewer challenge, and documented error reduction.
4. Advance responsibility only after the pass standard is demonstrated.
5. Use the final self-audit and mastery rubric for certification, while preserving the learner's error log and repeated-error rate.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 13

View saved version →

---
name: business-quality
description: Use for business-model deconstruction, unit economics, cohorts, pricing power, recurring revenue, retention, capital intensity, and reinvestment runway.
---

# Business Quality

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M026, M027, M028, M029, M030
- Appendices: None
- Master Labs: Master Lab 05
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Domain workflow

1. Read `references/operating-rules.md` and every bundled module reference that is material to the user's request.
2. Define the narrow decision question and the variables capable of changing value or risk.
3. Gather current primary evidence first when the task concerns a live public company.
4. Execute the explicit operating procedures in the loaded modules.
5. Preserve calculations, definitions, assumptions, contrary evidence, and source provenance.
6. Apply each loaded module's failure tests and completion test before calling the work complete.
7. Return a concise answer for a narrow question, but retain a reproducible evidence trail for every material claim.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 10

View saved version →

---
name: earnings-events
description: Use for earnings previews and post-earnings reviews, guidance math, estimate revisions, M&A, restructurings, activism, bankruptcy, distress, and event-driven company analysis.
---

# Earnings and Events

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M008, M009, M051, M052, M053, M054, M055
- Appendices: Appendix I
- Master Labs: Master Lab 09, Master Lab 10
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Event workflow

1. Freeze or reconstruct the pre-event expectation before interpreting the event.
2. Reconcile new reported facts to the relevant filing, earnings material, and footnotes.
3. Build the bridge from reported versus expected operating drivers through earnings, cash flow, financing, share count, and valuation.
4. Attribute every material estimate change to new evidence.
5. State what changed in the thesis, what did not change, and which future datapoint resolves the remaining debate.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 14

View saved version →

---
name: earnings-postmortem
description: Use to analyze earnings, attribute forecast errors, update the model, and record process lessons. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Earnings Postmortem

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M008-M009
- M051-M052
- M064-M065
- Appendix I

## Required workflow gates

1. reported versus expected bridge
2. forecast error attribution
3. definition changes
4. guidance and model update
5. thesis delta
6. process post-mortem

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: earnings-preview
description: Use to prepare a point-in-time earnings preview with expectations, scenarios, and decision-relevant watch items. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Earnings Preview

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M004-M005
- M051-M052
- M061
- Appendix I

## Required workflow gates

1. freeze cutoff
2. consensus and company guidance provenance
3. KPI bridge
4. scenario matrix
5. questions and signposts
6. post-event falsification plan

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: filings-primary-sources
description: Use for SEC filing retrieval, 10-K or 10-Q deep reads, 8-K triage, proxy research, filing comparison, XBRL reconciliation, and primary-source evidence work.
---

# Filings and Primary Sources

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M002, M004, M006, M007, M008, M009, M010
- Appendices: Appendix E
- Master Labs: Master Lab 08
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Domain workflow

1. Read `references/operating-rules.md` and every bundled module reference that is material to the user's request.
2. Define the narrow decision question and the variables capable of changing value or risk.
3. Gather current primary evidence first when the task concerns a live public company.
4. Execute the explicit operating procedures in the loaded modules.
5. Preserve calculations, definitions, assumptions, contrary evidence, and source provenance.
6. Apply each loaded module's failure tests and completion test before calling the work complete.
7. Return a concise answer for a narrow question, but retain a reproducible evidence trail for every material claim.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 13

View saved version →

---
name: financial-modeling
description: Use for historical normalization, revenue driver forecasts, margin forecasts, integrated balance sheet and cash flow modeling, scenarios, model architecture, and model QA.
---

# Financial Modeling

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M041, M042, M043, M044, M045
- Appendices: Appendix B, Appendix C
- Master Labs: Master Lab 02, Master Lab 14
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Modeling workflow

1. Read the operating rules and all bundled module references before building or changing a material model assumption.
2. Normalize historical definitions and periods before forecasting.
3. Forecast economic drivers first, then statements, cash, financing, and share count.
4. Use explicit scenario switches and preserve historical actuals.
5. Run deterministic calculations in `scripts/finance_engine.py` where applicable and reconcile results to sourced values.
6. Apply model-break tests from the references. Any hidden hard code, broken roll-forward, unexplained circularity, or non-reconciling statement is a failure.
7. Finish with assumptions, checks, sensitivities, model-change log, and unresolved modeling risks.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 14

View saved version →

---
name: forensic-accounting
description: Use for earnings-quality analysis, accrual tests, cash conversion, non-GAAP reconstruction, fraud red flags, management estimate bias, and forensic accounting diligence.
---

# Forensic Accounting

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M021, M022, M023, M024, M025
- Appendices: Appendix D
- Master Labs: Master Lab 04
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Domain workflow

1. Read `references/operating-rules.md` and every bundled module reference that is material to the user's request.
2. Define the narrow decision question and the variables capable of changing value or risk.
3. Gather current primary evidence first when the task concerns a live public company.
4. Execute the explicit operating procedures in the loaded modules.
5. Preserve calculations, definitions, assumptions, contrary evidence, and source provenance.
6. Apply each loaded module's failure tests and completion test before calling the work complete.
7. Return a concise answer for a narrow question, but retain a reproducible evidence trail for every material claim.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 12

View saved version →

---
name: forensic-screen
description: Use to run an accounting-forensics and earnings-quality screen before deeper underwriting. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Forensic Screen

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M011-M025
- Appendix D
- Master Lab 04

## Required workflow gates

1. source and period normalization
2. accruals and cash conversion
3. Beneish diagnostics
4. working-capital anomalies
5. non-GAAP and estimate review
6. disclosure drift
7. false-positive review

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: full-company-analysis
description: Use for full public-company initiations, deep stock research, maximum-depth underwriting, or requests to follow the complete institutional analyst manual end to end.
---

# Full Company Analysis

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M001, M002, M003, M004, M005, M006, M007, M008, M009, M010, M011, M012, M013, M014, M015, M016, M017, M018, M019, M020, M021, M022, M023, M024, M025, M026, M027, M028, M029, M030, M031, M032, M033, M034, M035, M036, M037, M038, M039, M040, M041, M042, M043, M044, M045, M046, M047, M048, M049, M050, M051, M052, M053, M054, M055, M056, M057, M058, M059, M060, M061, M062, M063, M064, M065, M066, M067, M068, M069, M070, M071, M072, M073, M074, M075, M076, M077, M078, M079, M080, M081, M082, M083, M084, M085, M086, M087, M088, M089, M090, M091, M092, M093, M094, M095, M096, M097, M098, M099, M100, M101, M102, M103, M104, M105, M106
- Appendices: Appendix A, Appendix B, Appendix C, Appendix D, Appendix E, Appendix F, Appendix G, Appendix H, Appendix I, Appendix J, Appendix K, Appendix L, Appendix M
- Master Labs: Master Lab 01, Master Lab 02, Master Lab 03, Master Lab 04, Master Lab 05, Master Lab 06, Master Lab 07, Master Lab 08, Master Lab 09, Master Lab 10, Master Lab 11, Master Lab 12, Master Lab 13, Master Lab 14
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Full workflow

1. Read `references/operating-rules.md`, `references/universal/front-matter-and-universal-standards.md`, and `references/module-router.json` first.
2. Define the decision question, security scope, valuation date, research horizon, and materiality threshold.
3. Build the source log and assumption register before drawing material conclusions.
4. Execute the Master 12-Step Research Workflow from the manual in sequence. Use the exact module references for each step rather than substituting a summary.
5. Identify the primary sector playbook and any genuinely material secondary playbook. Read `references/universal/sector-playbooks-common-protocol.md` before sector modules.
6. Use the applicable appendices and Master Labs for model build, valuation, forensics, risk, memo construction, and review.
7. For arithmetic, valuation, scenario aggregation, price-volume bridges, and liquidity calculations, prefer `scripts/finance_engine.py` when its inputs match the required definitions. Keep sourced inputs and units visible.
8. Maintain a two-layer output: a concise decision layer and a full evidence layer. Use `assets/full-research-output-template.md` as the deliverable skeleton unless the user asks for another format.
9. Run all applicable module failure tests and completion tests. Then run Appendix M. Record open gates in the completion audit.
10. Do not label the research complete when any mandatory gate remains open. Return the completed work plus a clearly separated unresolved-items section if public evidence is unavailable.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 151

View saved version →

---
name: full-company-initiation
description: Use to full institutional public-company initiation research and complete underwriting. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Full Company Initiation

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M001-M070
- primary M071-M106 sector playbook
- Appendices A-M
- applicable Master Labs 01-14

## Required workflow gates

1. scope and cutoff
2. mandatory live source set
3. historical reconstruction
4. driver model
5. valuation and reverse DCF
6. risk and red team
7. all audits
8. Appendix M

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: industry-competition
description: Use for industry structure, value chains, market sizing, TAM, competitive advantage, peer benchmarking, technology transitions, disruption, and competitive strategy.
---

# Industry and Competition

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M031, M032, M033, M034, M035
- Appendices: None
- Master Labs: Master Lab 05
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Domain workflow

1. Read `references/operating-rules.md` and every bundled module reference that is material to the user's request.
2. Define the narrow decision question and the variables capable of changing value or risk.
3. Gather current primary evidence first when the task concerns a live public company.
4. Execute the explicit operating procedures in the loaded modules.
5. Preserve calculations, definitions, assumptions, contrary evidence, and source provenance.
6. Apply each loaded module's failure tests and completion test before calling the work complete.
7. Return a concise answer for a narrow question, but retain a reproducible evidence trail for every material claim.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 11

View saved version →

---
name: industry-map
description: Use to map industry structure, value chain, market size, competition, technology transitions, and leading indicators. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Industry Map

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M031-M035
- Master Lab 05

## Required workflow gates

1. market boundary
2. value chain and profit pools
3. bottom-up TAM
4. competitor normalization
5. technology and substitution
6. leading indicators
7. falsification

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: management-governance
description: Use for management quality, capital allocation, executive incentives, insider transactions, board structure, governance risk, and management accountability.
---

# Management and Governance

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M036, M037, M038, M039, M040
- Appendices: Appendix K
- Master Labs: Master Lab 06
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Domain workflow

1. Read `references/operating-rules.md` and every bundled module reference that is material to the user's request.
2. Define the narrow decision question and the variables capable of changing value or risk.
3. Gather current primary evidence first when the task concerns a live public company.
4. Execute the explicit operating procedures in the loaded modules.
5. Preserve calculations, definitions, assumptions, contrary evidence, and source provenance.
6. Apply each loaded module's failure tests and completion test before calling the work complete.
7. Return a concise answer for a narrow question, but retain a reproducible evidence trail for every material claim.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 11

View saved version →

---
name: management-review
description: Use to evaluate management credibility, incentives, capital allocation, governance, and prediction history. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Management Review

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M010
- M036-M040
- Appendix K
- Master Lab 06

## Required workflow gates

1. prediction ledger
2. guidance accuracy
3. capital allocation scorecard
4. incentive and proxy review
5. insider and board evidence
6. credibility conclusion with counterevidence

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: peer-comparison
description: Use to build a definition-normalized operating, accounting, valuation, and risk peer comparison. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Peer Comparison

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M034
- M042
- M047
- sector playbooks

## Required workflow gates

1. peer rationale
2. definition register
3. accounting and calendar normalization
4. KPI comparison
5. valuation bridge
6. quality and risk adjustments
7. outliers and caveats

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: quarterly-update
description: Use to update an existing thesis, model, evidence graph, and valuation after a quarter. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Quarterly Update

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M008
- M042-M052
- M061-M068
- Appendix I

## Required workflow gates

1. load prior state
2. retrieve new filings and earnings materials
3. definition and restatement diff
4. forecast error attribution
5. model and valuation update
6. thesis delta
7. audits

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: research-communication
description: Use for investment memos, evidence design, charts and tables, investment committee defense, analyst communication, monitoring systems, and AI-assisted research controls.
---

# Research Communication

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M066, M067, M068, M069, M070
- Appendices: Appendix H, Appendix K
- Master Labs: Master Lab 10
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Communication workflow

1. Separate the decision layer from the evidence layer.
2. The decision layer should include the decision question, one-sentence thesis, variant perception, key operating drivers, valuation range, top risks, catalysts or validation timeline, thesis breaks, and what changed.
3. The evidence layer should preserve source log, definitions, historical reconstruction, model schedules, contrary evidence, risk register, valuation detail, and review history.
4. Do not make a persuasive narrative by deleting uncertainty or disconfirming evidence.
5. Apply the investment committee question bank and completion tests before finalizing a deep memo.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 12

View saved version →

---
name: reverse-dcf
description: Use to solve for market-implied operating expectations and test their plausibility. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Reverse Dcf

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M046-M050
- Appendix C
- Master Labs 03 and 14

## Required workflow gates

1. valuation-date EV bridge
2. clean cash-flow definition
3. solve implied growth or margin
4. translate into operational requirements
5. history and peer comparison
6. variant and falsification

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: risk-decision
description: Use for risk registers, balance-sheet stress, liquidity and covenant analysis, research sizing inputs, pre-mortems, decision journals, post-mortems, and error taxonomy.
---

# Risk and Decision Process

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M061, M062, M063, M064, M065
- Appendices: None
- Master Labs: Master Lab 09
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Risk workflow

1. Build risk chains, not a generic risk list. Each material risk needs an event, transmission mechanism, exposed metric, leading indicator, time horizon, severity or range, model line, mitigant, and thesis-break threshold.
2. Stress operating drivers first, then working capital, liquidity, covenants, refinancing, dilution, and valuation.
3. Use `scripts/finance_engine.py` for deterministic liquidity and scenario aggregation when definitions match.
4. Model correlated risks coherently rather than adding isolated worst cases.
5. Preserve a dated decision journal and state what future evidence changes the base, bear, stress, or thesis-break state.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 11

View saved version →

---
name: risk-stress-test
description: Use to build coherent operating, liquidity, covenant, refinancing, and dilution stress cases. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Risk Stress Test

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M050
- M055
- M061-M065
- Master Lab 09

## Required workflow gates

1. risk transmission chains
2. correlated scenario drivers
3. liquidity runway
4. debt maturities and covenants
5. refinancing and dilution
6. valuation damage
7. thesis breaks

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: sector-analysis
description: Use when research depends on sector-specific KPIs, accounting traps, primary sources, leading indicators, valuation methods, stress tests, or sector failure modes across the manual sector playbooks.
---

# Sector Analysis

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M071, M072, M073, M074, M075, M076, M077, M078, M079, M080, M081, M082, M083, M084, M085, M086, M087, M088, M089, M090, M091, M092, M093, M094, M095, M096, M097, M098, M099, M100, M101, M102, M103, M104, M105, M106
- Appendices: None
- Master Labs: Master Lab 11, Master Lab 13
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Sector workflow

1. Read `references/universal/sector-playbooks-common-protocol.md` first.
2. Select the sector playbook that best matches the company's economic engine, not merely its marketing label or exchange classification.
3. If a company spans sectors, identify one primary playbook and load secondary playbooks only for material segments.
4. Build the sector KPI history, definition register, leading-indicator dashboard, source map, accounting normalization, driver forecast, valuation, stress test, and exit-standard audit required by the chosen playbook.
5. Do not compare peer KPIs until definitions, geography, product mix, accounting, and capital structure are normalized.
6. End by testing the sector exit standard in the selected playbook.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 43

View saved version →

---
name: short-thesis-red-team
description: Use to challenge a proposed short thesis, identify squeeze and timing risks, and separate bad business from bad short. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Short Thesis Red Team

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M004
- M021-M025
- M031-M040
- M053-M055
- M061-M065

## Required workflow gates

1. borrow and catalyst assumptions
2. liquidity and solvency distinction
3. bull-case disconfirmation
4. crowding and timing
5. refinancing and strategic options
6. valuation asymmetry
7. thesis breaks

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: ten-k-deep-read
description: Use to perform a complete 10-K and footnote deep read with cross-filing comparisons. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Ten K Deep Read

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M006-M007
- M011-M025
- Appendices D-E

## Required workflow gates

1. document identity and amendments
2. statement and footnote map
3. accounting policy changes
4. commitments and contingencies
5. risk-factor and disclosure drift
6. numerical reconciliation
7. open questions

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: thesis-red-team
description: Use to independently challenge a long or neutral thesis without defending the base case. Produces a source-traceable workflow with deterministic calculations, explicit open gates, and an auditable decision layer.
---

# Thesis Red Team

Use this workflow only when it matches the user's requested scope. The original manual in ../../canonical is authoritative and must not be condensed away.

## Read first

- ../../references/v2/source-policy.md
- ../../references/v2/evidence-system.md
- ../../references/v2/state-and-gates.md
- ../../references/v2/audit-protocol.md
- ../../references/v2/confidence-scoring.md when scoring confidence
- ../../references/v2/research-budget-evoi.md for deep prioritization

## Canonical routing

- M004
- M021-M025
- M031-M040
- M061-M065
- Appendix D

## Required workflow gates

1. steelman thesis
2. attack evidence lineage
3. attack accounting and forecast assumptions
4. alternative causal model
5. liquidity and valuation vulnerabilities
6. ranked unresolved objections
7. required thesis response

Use the schemas in ../../schemas/v2, applicable deterministic scripts in ../../scripts, and the relevant sector model in ../../sector-models. For a current company, browsing and primary-source retrieval are mandatory. Classify each material statement, retain contrary evidence, and state the research cutoff. Run ../../scripts/research_audit.py before a completion claim. Failed mandatory gates require RESEARCH INCOMPLETE, with the missing evidence and decision impact named explicitly.

Do not execute securities transactions or present research confidence as a guarantee of investment outcome.

Referenced files: 1

View saved version →

---
name: valuation
description: Use for DCF, multiples, reverse DCF, market-implied expectations, sum-of-the-parts, valuation sensitivities, scenarios, and probability weighting.
---

# Valuation

Use this skill when the request matches the description above. Explicit user instructions determine scope and output format. Do not silently weaken source, calculation, falsification, or completion controls required by the loaded manual references.

## Mandatory references

- Modules: M046, M047, M048, M049, M050
- Appendices: Appendix C
- Master Labs: Master Lab 03, Master Lab 14
- Operating rules: `references/operating-rules.md`

The bundled DOCX is the canonical methodology. The Markdown files in `references/` are retrieval-oriented mirrors generated from that source. If an extracted reference appears malformed or ambiguous, consult the canonical DOCX when available rather than inventing missing content.

## Valuation workflow

1. Establish a clean valuation date, market capitalization or enterprise-value bridge, diluted share count, and capital structure.
2. Read all bundled valuation references before choosing methods.
3. Use operating-driver forecasts underneath intrinsic valuation. Normalize peer definitions before relative valuation.
4. Use at least two economically appropriate methods for deep work and include reverse expectations when public market price is available.
5. Run deterministic valuation math in `scripts/finance_engine.py` where applicable. Reconcile enterprise-to-equity adjustments and dilution.
6. Present scenario ranges, key sensitivities, market-implied assumptions, and the evidence that would falsify the valuation thesis.

## Evidence and output controls

- Distinguish reported facts, management claims, external estimates, analyst calculations, and analyst judgments.
- For current claims, verify freshness and state the as-of date.
- Cite material factual inputs to the actual source used.
- Recalculate important arithmetic rather than copying unexplained outputs.
- State unresolved questions and the strongest contrary case.
- Define specific falsification or thesis-break evidence for material conclusions.
- Do not execute securities transactions.

## V2 live-research and audit controls

For current-company work, read ../../references/v2/source-policy.md before research. Deep work must also use ../../references/v2/evidence-system.md, ../../references/v2/state-and-gates.md, and ../../references/v2/audit-protocol.md. Use the V2 schemas and deterministic scripts where applicable. Do not claim completion while a mandatory dependency, material contradiction, stale critical input, or unreconciled calculation remains open.

Referenced files: 13