What the skill tells Claude

This is the skill's instruction file, SKILL.md, as Claude reads it.

Purpose

Decompose a fixed income portfolio's active return versus its benchmark into four additive effects per bond:

Effect Formula What it measures
Income coupon_rate × (days / 365) Carry/coupon income
Treasury −ModDur × ΔYield_Treasury Interest rate sensitivity
Spread −ModDur × ΔSpread_Average_Class Sector credit risk
Specific Total_Return − Income − Treasury − Spread Idiosyncratic / issuer-specific

Invocation

/fi-attribution-campisi <user request>

Examples: - /fi-attribution-campisi Here is my fixed income portfolio xlsx - /fi-attribution-campisi Run Campisi attribution on the uploaded file

Input Schema

The skill accepts input in two forms:

Form 1 — Spec-compliant xlsx (preferred)

An Excel file containing three sheets: 1. Holdings: period_id, security_id, security_name, bond_class, rating_category, portfolio_weight, benchmark_weight, total_return, coupon_rate, modified_duration_start 2. Yield_Curve: tenor_years, rate_start, rate_end 3. Spread_Curves: bond_class, rating_category, avg_spread_start, avg_spread_end

Form 2 — Natural language + prompted collection

If no file is provided, the skill prompts for the required data.

Step-by-Step Workflow

The skill MUST execute these steps in order:

Step 1 — Detect input form

  1. Check whether the user uploaded an xlsx file.
  2. Otherwise, enter prompted collection mode.

Step 2 — Validate inputs (scripts/validate_inputs.py)

Run the validator against the detected input. The validator checks: - portfolio_weight sums to 1.0 ± 1e-4 - benchmark_weight sums to 1.0 ± 1e-4 - Required columns exist in Holdings, Yield_Curve, and Spread_Curves.

If validation passes → proceed to Step 3. If validation fails → display a clear error table and prompt the user to fix it.

Step 3 — Run attribution (scripts/run_attribution.py)

  1. Instantiate FixedIncomeCampisiAttribution from FixedIncomeCampisiAttribution.py
  2. Call generate_report() -> campisi_attribution_report.xlsx

Step 4 — Deliver results

Present a structured summary of the Active Effects and attach campisi_attribution_report.xlsx.

Prompt Rules (Prompted Collection Mode)

When inputs are missing or incomplete, prompt for them:

"I need the portfolio data to run the Campisi attribution. Please provide an Excel file with three sheets: 1. Holdings: Security-level data including weights, returns, coupon, duration, bond class, and rating. 2. Yield_Curve: Treasury yield curve rates at the start and end of the period. 3. Spread_Curves: Average OAS spreads by bond class and rating at the start and end of the period."

Files in This Skill Package

fi-attribution-campisi/
├── SKILL.md                        ← This file
├── assets/
│   └── icon.png                    ← Skill icon
└── scripts/
    ├── validate_inputs.py          ← Input validator
    ├── run_attribution.py          ← Attribution runner
    └── FixedIncomeCampisiAttribution.py ← Core engine

Shared as is, with no warranty. Not investment, legal or tax advice. See Disclosures.