BudgetTrackingSkill

---
name: BudgetTrackingSkill
description: >
  Compare planned budgets with actual spending, highlight significant gaps, and suggest corrective actions.
version: 1.0.1
visibility: team
tags:
  - finance
  - analytics
entrypoint:
  command: review_budget
  label: Review budget variance
---

## When Claude should load this skill
- Launch during monthly or quarterly variance analysis, finance business partner reviews, or budget vs. actual standups.
- Use when datasets contain cost centers, projects, or programs that require variance commentary and escalation guidance.

## Core workflow
1. Consult `instructions/budget-guidelines.md` for variance thresholds, commentary prompts, and approval routing.
2. Run `python scripts/gap_analysis.py` to emit `VARIANCE_SUMMARY` with variance totals and over-threshold alerts.
3. Draft management updates using `templates/variance-brief.md`, documenting mitigation actions for each owner.

## Linked resources
- `data/budget_plan.xlsx` — Planned spend by cost center, account, and month.
- `data/actual_spending.csv` — Actual transactions pulled from ERP or BI exports.
- `instructions/budget-guidelines.md` — Narrative guidance for interpreting favorable/unfavorable variances.
- `templates/variance-brief.md` — Markdown outline for presenting findings to budget owners.
- `data/risk_register.csv` — Optional list of high-risk programs requiring additional commentary.

import pandas as pd

THRESHOLD = 0.1

plan = pd.read_excel("data/budget_plan.xlsx")
actual = pd.read_csv("data/actual_spending.csv")
merged = plan.merge(actual, on=["cost_center", "month"], how="left")
merged["gap"] = (merged["actual"] - merged["plan"]) / merged["plan"]
alerts = merged[merged["gap"].abs() > THRESHOLD]

summary = {
    "total_variance": float(merged["gap"].sum()),
    "centers_over_threshold": alerts[["cost_center", "month", "gap"]].to_dict("records"),
}

print("VARIANCE_SUMMARY=", summary)