@v587d/capital-generation 2.7.6 → 2.7.8

package/selected-skills/financeskills/skills/audit-checklist/references/internal-controls.md

@@ -0,0 +1,36 @@
+# Internal Control Standards: COSO Framework
+
+The gold standard for designing, implementing, and assessing internal control for financial reporting.
+
+## 🏛️ The Five Components of COSO
+The COSO cube identifies five integrated components:
+
+### 1. Control Environment
+The set of standards, processes, and structures that provide the basis for carrying out internal control across the organization.
+- **Key Signal**: The "Tone at the Top" set by management.
+
+### 2. Risk Assessment
+A dynamic and iterative process for identifying and analyzing risks to achieving objectives.
+- **Key Signal**: Regular SWOT analyses and risk registries.
+
+### 3. Control Activities
+The actions established through policies and procedures that help ensure that management’s directives to mitigate risks to the achievement of objectives are carried out.
+- **Examples**: Approvals, authorizations, verifications, reconciliations, and segregation of duties.
+
+### 4. Information & Communication
+The internal and external communication of information necessary to support the functioning of other components of internal control.
+
+### 5. Monitoring Activities
+Ongoing evaluations, separate evaluations, or some combination of the two used to ascertain whether each of the five components of internal control is present and functioning.
+
+---
+
+## 🔍 Common Control Gaps
+- **Lack of Segregation of Duties**: The same person can initiate, approve, and record a transaction.
+- **Inadequate Documentation**: No audit trail for approvals.
+- **Override of Controls**: Management bypassing established procedures.
+
+## 🔗 Sources
+- [COSO Official Website](https://www.coso.org/)
+- [AICPA: Internal Control - Integrated Framework](https://www.aicpa.org/resources/toolkit/coso)
+- [Deloitte: COSO Internal Control Integrated Framework](https://www2.deloitte.com/us/en/pages/advisory/articles/internal-control-integrated-framework.html)

package/selected-skills/financeskills/skills/audit-checklist/references/isa-standards.md

@@ -0,0 +1,26 @@
+# International Standards on Auditing (ISA)
+
+Overview of the professional standards for the performance of financial audits of historical financial information.
+
+## 🏛️ Core Principles
+
+### ISA 200: Overall Objectives
+The objective of an audit is to obtain reasonable assurance about whether the financial statements as a whole are free from material misstatement, whether due to fraud or error.
+
+### ISA 240: Auditor’s Responsibilities Relating to Fraud
+Distinguishes between fraud (deliberate) and error (unintentional). Requires auditors to maintain professional skepticism throughout the audit.
+
+### ISA 315: Identifying Risks of Material Misstatement
+The auditor must understand the entity and its environment, including internal controls, to identify and assess the risks of material misstatement.
+
+---
+
+## 🔍 The Audit Opinion Types
+1.  **Unmodified (Unqualified)**: The "Clean" opinion. Statements are presented fairly.
+2.  **Qualified**: Material but not pervasive misstatement found. "Except for [X], everything is fair."
+3.  **Adverse**: Material and pervasive misstatement. The statements are misleading.
+4.  **Disclaimer**: The auditor is unable to obtain sufficient evidence to form an opinion.
+
+## 🔗 Sources
+- [IAASB Standards & Pronouncements](https://www.iaasb.org/standards-pronouncements)
+- [IFAC: International Standards on Auditing](https://www.ifac.org/what-we-do/standards/auditing-assurance)

package/selected-skills/financeskills/skills/automated-reconciliation/EXAMPLE.md

@@ -0,0 +1,21 @@
+# Worked Example - automated-reconciliation
+
+## Ask
+
+> I have 500 bank transactions and 510 ledger entries. Most are exact matches, but 20 have slightly different names like 'AWS' vs 'Amazon'. How can I reconcile this?
+
+## What a correct response contains
+
+Should recommend a two-step process: Deterministic matching for exact entries and Probabilistic/Fuzzy matching for name variances. Should mention vendor name normalization. Should provide a Reconciliation Report structure.
+
+## File-driven variant
+
+With fixture(s) `evals/files/bank_statement.csv`, `evals/files/gl_cash.csv`:
+
+> Reconcile the attached May bank statement (evals/files/bank_statement.csv) against the GL cash ledger (evals/files/gl_cash.csv). Both start from an agreed opening balance of $100,000.00. Match transactions, identify deposits in transit, outstanding cheques, and bank-only items, then produce a bank reconciliation statement proving the adjusted balances agree.
+
+Expected: Bank ending balance $147,060.00; GL ending balance $154,540.00. Deposit in transit $27,300.00 (INV1045); outstanding cheques $12,640.00 (2206) and $7,120.00 (2207); bank-only items: charges $350.00 and interest $410.00 to be booked in GL. Adjusted balance both sides $154,600.00.
+
+## Compute, don't estimate
+
+All figures above come from [`scripts/calculate.py`](./scripts/calculate.py) - import its functions rather than doing arithmetic in prose.

package/selected-skills/financeskills/skills/automated-reconciliation/SKILL.md

@@ -0,0 +1,87 @@
+---
+name: automated-reconciliation
+description: When the user wants to match disparate financial data sources (e.g., bank statements vs. general ledger) at scale using AI. Also use when the user mentions "bank rec," "matching invoices," "reconciling payments," "fuzzy matching in finance," or "automated accounting."
+metadata:
+  version: 1.0.0
+---
+
+# Automated Reconciliation
+
+You are an Accounting Systems Architect. Your goal is to eliminate manual data entry by using fuzzy matching and AI to reconcile thousands of transactions in seconds.
+
+## Initial Assessment
+
+1. **The Data Sources**
+   - Source A: (e.g., Bank Statement PDF/CSV).
+   - Source B: (e.g., General Ledger / ERP Export).
+   - Are there common IDs (Reference numbers, Check numbers)?
+
+2. **The "Match" Definition**
+   - Exact match (Same ID, Same Amount, Same Date).
+   - Fuzzy match (Similar Name, Same Amount, +/- 2 days).
+
+---
+
+## Reconciliation Framework
+
+### Technical Limitation
+**LLMs are not good at matching 50,000 rows.**
+For large datasets, this skill uses Python libraries like `pandas` and `RecordLinkage`. LLMs are used to resolve the "Ambiguous Matches" (the 5% the code can't solve).
+
+### Priority Order
+1. **Data Cleaning** (Standardizing vendor names: "AWS" vs. "Amazon Web Svcs").
+2. **Deterministic Matching** (Exact matches on IDs).
+3. **Probabilistic (Fuzzy) Matching** (Using Jaro-Winkler or Levenshtein distance).
+4. **Exception Handling** (Flagging the items that couldn't be matched).
+
+---
+
+## Technical Reconciliation Steps
+
+### 1. Vendor Name Normalization
+- Use a lookup table or AI to standardize variant vendor names across datasets.
+
+### 2. Fuzzy Amount Matching
+- Use a tolerance window (e.g., match if amounts are within $0.05 to account for rounding errors).
+
+### 3. Many-to-One Resolution
+- Identify cases where one bank deposit represents three separate invoices in the ledger.
+
+---
+
+## Output Format
+
+### Reconciliation Report Structure
+
+**The Results**
+- **Match Rate**: (e.g., 94% matched automatically).
+- **Total Reconciled Value**: $X.
+
+**The Exceptions**
+- List of "Unmatched" items from both sources.
+- List of "Ambiguous" matches requiring human sign-off (with confidence scores).
+
+**Journal Entry Suggestions**
+- Ready-to-copy entries for bank fees or interest detected in the statement but missing from the ledger.
+
+---
+
+## Scripts
+- [calculate.py](./scripts/calculate.py): Exact, tolerance, and subset-sum matching engine. Run with `python3 scripts/calculate.py` to self-test; import the functions for actual computations.
+
+---
+
+## References
+- [Fuzzy Matching for Finance](./references/fuzzy-logic.md): Jaro-Winkler vs. Levenshtein.
+- [Rec Best Practices](./references/rec-controls.md): Maintaining an audit trail.
+
+---
+
+## Assets
+- [reconciliation-statement-template.md](./assets/reconciliation-statement-template.md): Bank rec statement with outstanding-items detail.
+
+---
+
+## Related Skills
+- **financial-statement-prep**: To ensure the cash balance on the balance sheet is accurate.
+- **audit-checklist**: For auditing the reconciliation process itself.

package/selected-skills/financeskills/skills/automated-reconciliation/assets/reconciliation-statement-template.md

@@ -0,0 +1,31 @@
+# Bank Reconciliation Statement
+
+**Account:** {bank, account no} | **Period:** {month} | **Prepared:** {date} | **Preparer:** AI agent (review required)
+
+## Reconciliation
+| | Amount |
+|---|---|
+| Balance per bank statement, {date} | {x} |
+| Add: deposits in transit | {x} |
+| Less: outstanding cheques/payments | ({x}) |
+| **Adjusted bank balance** | **{x}** |
+| Balance per general ledger, {date} | {x} |
+| Add: bank credits not booked (interest, collections) | {x} |
+| Less: bank debits not booked (charges, returned items) | ({x}) |
+| **Adjusted GL balance** | **{x}** |
+
+Difference: {must be 0.00 - if not, list unresolved items below}
+
+## Outstanding Items Detail
+| Type | Date | Ref | Description | Amount | Age (days) |
+|---|---|---|---|---|---|
+
+## Adjusting Journal Entries Required
+| Account | Dr | Cr | Memo |
+|---|---|---|---|
+
+## Exceptions / Unresolved
+{Items needing investigation, with owner and due date. Flag anything > 30 days old.}
+
+---
+*Matching performed with scripts/calculate.py. Reviewed by: ____________ Date: ________*

package/selected-skills/financeskills/skills/automated-reconciliation/references/fuzzy-logic.md

@@ -0,0 +1,27 @@
+# Fuzzy Matching Logic for Reconciliations
+
+## Matching Tiers (run in order, lock matched items after each pass)
+1. **Exact**: amount + date + reference all equal.
+2. **Tolerance**: amount within ±0.5% or fixed threshold (e.g., ₦100/$1 for bank fees), date within ±3 business days, reference exact.
+3. **Fuzzy reference**: amount within tolerance + string similarity on description/reference ≥ threshold.
+4. **One-to-many / many-to-many**: subset-sum matching (one GL entry vs. several bank lines, e.g., batched settlements). Cap combination size at 5 to control false positives.
+
+## String Similarity Techniques
+- **Normalize first**: uppercase, strip punctuation/whitespace, remove noise tokens ("PAYMENT", "TRF", "REF"), expand known abbreviations from a maintained alias table.
+- **Levenshtein ratio**: good for typos; threshold ≥ 0.85.
+- **Token-set ratio** (rapidfuzz): robust to word reordering ("ACME LTD INVOICE 123" vs "INV123 ACME LIMITED").
+- **Vendor alias map**: persist confirmed matches into a lookup so the same counterparty never needs fuzzy matching twice.
+
+## Python Sketch
+```python
+from rapidfuzz import fuzz
+def is_match(a, b, amt_a, amt_b, tol=0.005):
+    amt_ok = abs(amt_a - amt_b) <= tol * max(abs(amt_a), abs(amt_b))
+    return amt_ok and fuzz.token_set_ratio(a, b) >= 85
+```
+
+## Guardrails
+- Never auto-match across different currencies without FX conversion at transaction-date rate.
+- Flag fuzzy matches ≥ materiality for human review; auto-clear only below threshold.
+- Log match tier + score on every pairing — auditors will ask.
+- False-positive monitoring: sample 25 auto-matches monthly; > 2 errors → raise thresholds.

package/selected-skills/financeskills/skills/automated-reconciliation/references/rec-controls.md

@@ -0,0 +1,28 @@
+# Reconciliation Controls and Standards
+
+## Control Attributes of a Complete Reconciliation
+1. **Preparer and reviewer are different people**; reviewer is senior; both sign and date.
+2. **Timeliness**: completed within close calendar deadline (typically WD3–WD5).
+3. **Source independence**: GL balance vs. independent source (bank statement, subledger, custodian statement, counterparty confirmation).
+4. **Reconciling items**: each item aged, explained, owner-assigned, with clearance date. "Unknown difference" is not an explanation.
+5. **Materiality threshold** documented; unreconciled residual below threshold still listed, never silently plugged.
+
+## Reconciling Item Categories
+| Category | Example | Action |
+|---|---|---|
+| Timing | Deposit in transit, outstanding check | Self-clears; monitor aging |
+| Error — ours | Transposition, wrong account | Correcting JE this period |
+| Error — theirs | Bank error | Dispute with bank, track to resolution |
+| Unidentified | Unmatched residual | Investigate; escalate if > threshold or > 30 days |
+
+## Aging and Escalation Policy (typical)
+- Items > 30 days: escalate to controller.
+- Items > 60 days: written explanation to CFO.
+- Items > 90 days: mandatory resolution or P&L write-off with approval.
+
+## Red Flags (fraud-relevant)
+- Recurring round-amount reconciling items.
+- The same item re-explained differently across months.
+- Reconciliations done but never reviewed (check reviewer timestamps).
+- Suspense/clearing accounts used as permanent parking.
+- Preparer with sole access to both the GL and the source system (SoD breach).

package/selected-skills/financeskills/skills/banking-compliance/EXAMPLE.md

@@ -0,0 +1,21 @@
+# Worked Example - banking-compliance
+
+## Ask
+
+> Our bank has $100M in CET1 capital and $800M in Risk-Weighted Assets. Are we Basel III compliant?
+
+## What a correct response contains
+
+Should calculate the CET1 ratio ($100M / $800M = 12.5%). Should compare this to the Basel III minimum (4.5% + buffers). Should conclude the bank is currently compliant. Should provide a Regulatory Compliance Dashboard.
+
+## File-driven variant
+
+With fixture(s) `evals/files/exposures.csv`:
+
+> CET1 capital is $410,000,000 and the countercyclical buffer is 1%. Using the attached exposure file (evals/files/exposures.csv), compute RWA under the standardised approach, the CET1 ratio, and assess compliance against minimum 4.5% + 2.5% conservation + 1% CCyB.
+
+Expected: RWA $3,645,000,000; CET1 ratio 11.25% vs requirement 8.0%; surplus 3.25% - compliant.
+
+## Compute, don't estimate
+
+All figures above come from [`scripts/calculate.py`](./scripts/calculate.py) - import its functions rather than doing arithmetic in prose.

package/selected-skills/financeskills/skills/banking-compliance/SKILL.md

@@ -0,0 +1,82 @@
+---
+name: banking-compliance
+description: When the user wants to analyze bank capital adequacy or regulatory compliance (Basel III). Also use when the user mentions "Common Equity Tier 1 (CET1)," "RWA," "Risk-Weighted Assets," "liquidity coverage ratio," "leverage ratio," or "bank stress testing."
+metadata:
+  version: 1.0.0
+---
+
+# Banking Regulatory Compliance (Basel III)
+
+You are a Bank Risk & Compliance Officer. Your goal is to ensure the bank maintains sufficient capital and liquidity buffers to survive economic shocks.
+
+## Initial Assessment
+
+1. **Capital Components**
+   - **Common Equity Tier 1 (CET1)**: Common shares + Retained earnings.
+   - **Tier 1 Capital**: CET1 + Additional Tier 1.
+   - **Total Capital**: Tier 1 + Tier 2.
+
+2. **Asset Risk Profile**
+   - What are the **Risk-Weighted Assets (RWA)**? (e.g., Corporate loans carry higher weight than Govt bonds).
+
+3. **Liquidity Markers**
+   - **LCR (Liquidity Coverage Ratio)**: High-quality liquid assets / Net cash outflows over 30 days.
+
+---
+
+## Compliance Framework
+
+### Priority Order
+1. **CET1 Ratio Calculation** (`CET1 / RWA`).
+2. **Tier 1 & Total Capital Adequacy**.
+3. **Liquidity Buffer Review** (LCR & NSFR).
+4. **Leverage Ratio Assessment** (Tier 1 Capital / Total Exposure).
+5. **Stress Test Modeling** (Impact of credit defaults on capital ratios).
+
+---
+
+## Technical Compliance Steps
+
+### 1. Risk-Weighting
+- Assign 0% weight to cash/sovereigns.
+- Assign 35-100% to mortgages.
+- Assign 100%+ to corporate loans.
+
+### 2. Capital Buffers
+- Calculate the Capital Conservation Buffer (CCB) and Countercyclical Buffer (CCyB).
+
+---
+
+## Output Format
+
+### Regulatory Compliance Dashboard
+
+**Capital Adequacy**
+- **CET1 Ratio**: (Target > 4.5% + Buffers).
+- **Tier 1 Ratio**: (Target > 6.0%).
+- **Total Capital Ratio**: (Target > 8.0%).
+
+**Liquidity Position**
+- **LCR**: (Target > 100%).
+- **NSFR**: (Net Stable Funding Ratio).
+
+**Compliance Status**
+- Green/Yellow/Red status for each regulatory threshold.
+
+---
+
+## Scripts
+- [calculate.py](./scripts/calculate.py): Deterministic functions for this skill's core computations, kept as reviewable material. This plugin's research role has no shell and the scripts are not part of the installed package, so nothing here runs them: work the ratios from [Basel III Standards](./references/basel-iii-framework.md) and the weight bands in [RWA Classification](./references/risk-weights.md), show each calculation in the answer, and hand anything that needs real compute to the delegated `data_junior`.
+
+---
+
+## References
+- [Basel III Standards](./references/basel-iii-framework.md): The official BIS guidelines.
+- [RWA Classification](./references/risk-weights.md): How to weight different bank assets.
+
+---
+
+## Related Skills
+- **ecl-computation**: Basel III capital is the buffer against ECL losses.
+- **risk-assessment**: For evaluating the market and credit risks driving RWA.
+- **corporate-consolidation**: For group-level capital adequacy reporting.

package/selected-skills/financeskills/skills/banking-compliance/references/basel-iii-framework.md

@@ -0,0 +1,26 @@
+# Basel III: Regulatory Capital Framework
+
+The global regulatory framework for bank capital adequacy, stress testing, and market liquidity risk.
+
+## 🏛️ The Three Pillars
+
+### Pillar 1: Minimum Capital Requirements
+- **CET1 Ratio**: Minimum 4.5% of Risk-Weighted Assets (RWA).
+- **Tier 1 Ratio**: Minimum 6.0% of RWA.
+- **Total Capital Ratio**: Minimum 8.0% of RWA.
+
+### Pillar 2: Supervisory Review
+Ensures that banks have adequate internal processes to assess their capital adequacy and that supervisors review these assessments.
+
+### Pillar 3: Market Discipline
+Mandatory disclosure requirements to provide market participants with information about a bank's risk profile and capital adequacy.
+
+---
+
+## 🔍 Key Buffers
+- **Capital Conservation Buffer (CCB)**: An additional 2.5% of CET1 to withstand periods of stress.
+- **Countercyclical Buffer (CCyB)**: Applied during periods of high credit growth to protect the banking sector.
+
+## 🔗 Sources
+- [Bank for International Settlements (BIS): Basel III](https://www.bis.org/bcbs/basel3.htm)
+- [Investopedia: Basel III](https://www.investopedia.com/terms/b/basel-iii.asp)

package/selected-skills/financeskills/skills/banking-compliance/references/risk-weights.md

@@ -0,0 +1,32 @@
+# Basel III Standardized Risk Weights (Credit Risk)
+
+## Common Exposure Classes (Standardized Approach, Basel III final)
+| Exposure | Risk Weight |
+|---|---|
+| Sovereigns AAA to AA- | 0% |
+| Sovereigns A+ to A- | 20% |
+| Sovereigns BB+ to B- | 100% |
+| Banks (external rating A) | 30%–50% (by grade/tenor) |
+| Corporates AAA to AA- | 20% |
+| Corporates A | 50% |
+| Corporates BBB | 75%–100% |
+| Corporates unrated | 100% (85% if SME under some regimes) |
+| Regulatory retail (qualifying) | 75% |
+| Residential mortgage | 20%–70% by LTV band |
+| Commercial real estate | 60%–110% by LTV/repayment dependence |
+| Defaulted exposures | 100%–150% |
+| Equity | 250%–400% |
+
+National regulators (e.g., CBN) adapt these — always check the local prudential guideline; CBN applies its own weights for FGN securities and state government exposures.
+
+## Capital Ratios
+- `CET1 ratio = CET1 / RWA ≥ 4.5%` (+ capital conservation buffer 2.5% + D-SIB/G-SIB surcharges + countercyclical buffer).
+- Tier 1 ≥ 6%, Total capital ≥ 8%. CBN minimum CAR: 10% (national), 15% (international authorization).
+
+## Off-Balance Sheet
+`Credit equivalent = Notional × CCF` — direct credit substitutes 100%, performance bonds 50%, self-liquidating trade LCs 20%, unconditionally cancellable commitments 10%.
+
+## Other RWA Components
+- Operational risk: Standardized Measurement Approach (BI component × marginal coefficients × ILM).
+- Market risk: FRTB standardized sensitivities-based method.
+- Output floor: aggregate IRB RWA ≥ 72.5% of standardized.

package/selected-skills/financeskills/skills/budget-forecast/EXAMPLE.md

@@ -0,0 +1,21 @@
+# Worked Example - budget-forecast
+
+## Ask
+
+> We need a budget for 2025. Last year revenue was $10M with 15% growth. We want to double our marketing spend to $1M. Can you build a high-level roadmap?
+
+## What a correct response contains
+
+Should identify revenue drivers (base revenue + growth rate). Should calculate projected revenue ($11.5M). Should account for fixed vs variable costs. Should discuss the impact of doubling marketing spend on EBITDA. Should provide a structured plan with Executive Summary and Scenario Analysis.
+
+## File-driven variant
+
+With fixture(s) `evals/files/history_and_drivers.csv`:
+
+> From the attached history and drivers (evals/files/history_and_drivers.csv): compute the 2023-2025 revenue CAGR, then build the 2026 driver-based budget (units x price, variable % of revenue, fixed costs) and state budgeted operating income. Sanity-check the implied 2026 growth against the historical CAGR.
+
+Expected: CAGR 31.5%; 2026 budget revenue $66,000,000, operating income $11,700,000; implied growth 14.6% vs CAGR 31.5%.
+
+## Compute, don't estimate
+
+All figures above come from [`scripts/calculate.py`](./scripts/calculate.py) - import its functions rather than doing arithmetic in prose.

package/selected-skills/financeskills/skills/budget-forecast/SKILL.md

@@ -0,0 +1,87 @@
+---
+name: budget-forecast
+description: When the user wants to create a budget, forecast future performance, or perform variance analysis. Also use when the user mentions "financial projections," "2025 planning," "burn rate forecast," "revenue modeling," "budget vs actual," "planning for next year," or "financial roadmap." Use this for operational planning and strategic forecasting.
+metadata:
+  version: 1.1.0
+---
+
+# Budget Forecasting
+
+You are a Financial Planning & Analysis (FP&A) Manager. Your goal is to build accurate, data-driven projections that help the business allocate resources effectively.
+
+## Initial Assessment
+
+1. **Forecasting Horizon**
+   - Short-term (monthly), Medium-term (annual), or Long-term (3-5 years)?
+   - What is the specific business unit or product line?
+
+2. **Driver Identification**
+   - What are the primary revenue drivers? (Units, Price, Retention).
+   - What are the primary cost drivers? (Headcount, Marketing, COGS).
+
+3. **Historical Context**
+   - Do we have 2-3 years of historical Actuals to build from?
+   - Any major expected shifts in strategy or market?
+
+---
+
+## Forecasting Framework
+
+### Priority Order
+1. **Revenue Modeling** (Top-line growth assumptions).
+2. **Resource Allocation** (Headcount and OpEx planning).
+3. **Cash Flow Impact** (Timing of inflows/outflows).
+4. **Scenario Analysis** (Best, Worst, and Base cases).
+5. **Variance Tracking** (Setting up the Budget vs. Actual framework).
+
+---
+
+## Technical Modeling Steps
+
+### 1. Revenue Drivers
+- Build a bottom-up model based on sales pipeline, churn rates, and average contract value (ACV).
+- Apply seasonal adjustments based on historical trends.
+
+### 2. Expense Planning
+- **Fixed Costs**: Rent, salaries, depreciation.
+- **Variable Costs**: Commissions, cloud hosting, variable marketing.
+
+### 3. Sensitivity Analysis
+- "What-if" modeling: "If churn increases by 1%, what happens to our end-of-year cash balance?"
+
+---
+
+## Output Format
+
+### Financial Plan Structure
+
+**Executive Summary**
+- Key Financial Targets (Revenue, EBITDA, Net Cash Flow).
+- Major assumptions list.
+
+**The Projections**
+- Monthly/Quarterly P&L forecast.
+- Personnel/Headcount plan.
+- Capital Expenditure (CapEx) roadmap.
+
+**Risk & Scenarios**
+- Scenario matrix (Bull vs. Bear).
+- Sensitivity table for primary drivers.
+
+---
+
+## Scripts
+- [calculate.py](./scripts/calculate.py): Deterministic functions for this skill's core computations. Run `python3 scripts/calculate.py` to self-test; import the functions instead of doing mental math.
+
+---
+
+## References
+- [Forecasting Best Practices](./references/forecasting-best-practices.md): Evidence-backed modeling standards.
+- [Scenario Analysis Guide](./references/scenario-analysis.md): How to structure what-if models.
+
+---
+
+## Related Skills
+- **financial-analysis**: For auditing the 'Actuals' part of Budget vs. Actual.
+- **risk-assessment**: For evaluating the risks inherent in the forecast assumptions.
+- **investment-analysis**: For using the forecast to value the entity.

package/selected-skills/financeskills/skills/budget-forecast/references/forecasting-best-practices.md

@@ -0,0 +1,26 @@
+# Forecasting Best Practices
+
+Building accurate financial projections requires a blend of historical data, market insight, and statistical discipline.
+
+## 🏛️ Core Principles
+
+### 1. Driver-Based Modeling
+Identify the "Levers" that control the numbers. Instead of just guessing revenue, model it as:
+`Traffic x Conversion Rate x Average Order Value = Revenue`.
+
+### 2. Rolling Forecasts
+Don't just set a budget once a year. Update the forecast monthly or quarterly as actual data comes in (e.g., 4+8, 9+3 reporting).
+
+### 3. Variance Analysis (BvA)
+Regularly compare Budget vs. Actual. Analyze the **Price Variance** (we sold it for less than expected) and **Volume Variance** (we sold fewer units than expected).
+
+---
+
+## 🔍 Common Pitfalls
+- **Over-Optimism Bias**: Underestimating the time or cost to achieve growth.
+- **Sandbagging**: Intentionally lowering targets to ensure they are beaten.
+- **Ignoring Seasonality**: Assuming a straight-line growth in businesses with holiday peaks.
+
+## 🔗 Sources
+- [CFI: Financial Forecasting Guide](https://corporatefinanceinstitute.com/resources/fp-a/financial-forecasting-guide/)
+- [FP&A Trends: Best Practices in Financial Forecasting](https://fpa-trends.com/)

package/selected-skills/financeskills/skills/budget-forecast/references/scenario-analysis.md

@@ -0,0 +1,25 @@
+# Scenario Analysis: Modeling the Unknown
+
+How to structure "What-If" scenarios to test the resilience of a financial plan.
+
+## 🏛️ Types of Scenarios
+
+### 1. The Base Case
+The most likely outcome based on current trends and management's stated strategy.
+
+### 2. The Bull Case (Upside)
+What happens if everything goes right? (e.g., 2x market growth, successful product launch).
+
+### 3. The Bear Case (Downside)
+What happens if things go wrong? (e.g., major competitor enters, recession, supply chain failure).
+
+---
+
+## 🔍 How to Build a Scenario Matrix
+1.  **Identify Key Variables**: Select the 2-3 variables that have the highest impact on EBITDA (e.g., Sales Volume, Labor Cost).
+2.  **Define Thresholds**: Use +/- 10% or historical worst-case data points.
+3.  **Calculate Outcomes**: Run the model with these inputs to see the impact on cash flow and covenant compliance.
+
+## 🔗 Sources
+- [Investopedia: Scenario Analysis](https://www.investopedia.com/terms/s/scenario_analysis.asp)
+- [Harvard Business Review: Why You Need to Model Multiple Futures](https://hbr.org/2020/07/why-you-need-to-model-multiple-futures)

package/selected-skills/financeskills/skills/close-management/EXAMPLE.md

@@ -0,0 +1,17 @@
+# Worked Example - close-management
+
+## Ask
+
+> Our close takes 9 working days. Subledgers close WD2, recs finish WD6 because accruals aren't done until WD4, and consolidation waits for one entity that always locks late. Build us a plan to get to 5 days.
+
+## What a correct response contains
+
+Should identify the critical path (accruals -> recs -> entity lock -> consolidation), move predictable accruals pre-close, set intercompany/late-entity deadlines with escalation, and produce a day-by-day close calendar with dependencies.
+
+## File-driven variant
+
+With fixture(s) `evals/files/close_status_day4.csv`:
+
+> It is Day 4 of the close. Assess the attached status (evals/files/close_status_day4.csv): identify the critical path to sign-off, what is late, the blocker to escalate, and a recovery plan that still hits Day 5.
+
+Expected: Critical path T-03/T-04 -> T-06 -> T-07. T-03 is 2 days late with no dependency - start immediately. T-04 blocked on SubB - escalate today; if statement not received by EOD, eliminate using SubB's prior balance + accrual and true-up next month. T-06 compresses to Day 5 morning; T-07 Day 5 afternoon.