expense-categorization
Classify expenses by category, department, and tax deductibility from transaction data. Use when the user requests expense categorization or provides relevant inputs for this workflow.
git clone --depth 1 https://github.com/seb1n/awesome-ai-agent-skills /tmp/expense-categorization && cp -r /tmp/expense-categorization/finance-and-accounting/expense-categorization ~/.claude/skills/expense-categorizationSKILL.md
# Expense Categorization Automatically classify business expenses into accounting categories, assign department cost centers, and flag tax-deductible items from raw transaction data. This skill processes credit card statements, bank feeds, and expense reports to produce clean, categorized output suitable for bookkeeping, tax preparation, and spend analytics. ## Workflow 1. **Receive Expense Data** Accept transaction data as CSV, bank statement text, or structured records. Required fields: date, amount, and description or merchant name. Optional fields: card last four digits, employee name, department, receipt notes. Normalize date formats and currency to a consistent standard. 2. **Parse Description and Merchant** Extract the merchant name from the transaction descriptor, stripping out authorization codes, location suffixes, and card network prefixes. Map common merchant name variations to canonical names (e.g., "AMZN MKTP US" → "Amazon", "GOOGLE *GSUITE" → "Google Workspace"). Use the merchant category code (MCC) when available as a secondary signal. 3. **Classify Expense Category** Assign each transaction to a primary category based on merchant identity, MCC, description keywords, and amount patterns. Standard categories: Travel & Lodging, Meals & Entertainment, Software & SaaS, Office Supplies, Professional Services, Advertising, Utilities, Insurance, Shipping & Postage, Equipment, Training & Education, Miscellaneous. 4. **Assign Department and Cost Center** Route each expense to the appropriate department based on the cardholder, project codes in the description, or pre-configured rules. Apply default department assignments for known merchants (e.g., AWS charges → Engineering, HubSpot → Marketing). 5. **Flag Tax-Deductible Items** Mark expenses that qualify for tax deduction based on category and business purpose. Apply IRS rules for meals (50% deductible for business meals, 100% for company events), home office, vehicle mileage, and professional development. Flag items that need substantiation — receipts, business purpose memo, or attendee lists. 6. **Generate Categorized Summary** Output a categorized transaction list with totals by category, department, and tax status. Highlight any transactions that couldn't be confidently categorized (confidence < 80%) for manual review. ## Usage Provide transaction data in any tabular format. Specify any custom category mappings, department rules, or tax jurisdiction if not US federal. **Example prompt:** > Categorize these November credit card transactions for our 15-person startup. We have Engineering, Marketing, and Operations departments. Flag tax-deductible items. Here's the CSV: [paste transactions] ## Examples ### Example 1: Monthly Credit Card Categorization **Input (CSV):** ```csv date,amount,description,cardholder 2024-11-02,249.00,GITHUB TEAM PLAN,Sarah Chen 2024-11-03,1450.00,UNITED AIRLINES 0167823,Mike Torres 2024-11-03,189.50,MARRIOTT HOTEL CHI,Mike Torres 2024-11-05,42.67,UBER TRIP FKJE83,Mike Torres 2024-11-08,156.00,GOOGLE ADS 8834721,Lisa Park 2024-11-12,89.99,ZOOM VIDEO COMMS,Sarah Chen 2024-11-14,67.32,DOORDASH DASHER TIP,Operations 2024-11-18,4200.00,DELL TECHNOLOGIES,Sarah Chen 2024-11-22,350.00,COURSERA BUSINESS,Mike Torres 2024-11-25,23.45,STAPLES #1284,Operations ``` **Output:** | Date | Amount | Merchant | Category | Department | Tax Deductible | Notes | |------------|-----------|------------------|-----------------------|-------------|----------------|-------------------------------| | 2024-11-02 | $249.00 | GitHub | Software & SaaS | Engineering | Yes (100%) | | | 2024-11-03 | $1,450.00 | United Airlines | Travel & Lodging | Engineering | Yes (100%) | Business travel | | 2024-11-03 | $189.50 | Marriott | Travel & Lodging | Engineering | Yes (100%) | Business travel | | 2024-11-05 | $42.67 | Uber | Travel & Lodging | Engineering | Yes (100%) | Ground transport | | 2024-11-08 | $156.00 | Google Ads | Advertising | Marketing | Yes (100%) | | | 2024-11-12 | $89.99 | Zoom | Software & SaaS | Engineering | Yes (100%) | | | 2024-11-14 | $67.32 | DoorDash | Meals & Entertainment | Operations | Yes (50%) | Needs business purpose memo | | 2024-11-18 | $4,200.00 | Dell | Equipment | Engineering | Depreciation | Section 179 eligible | | 2024-11-22 | $350.00 | Coursera | Training & Education | Engineering | Yes (100%) | | | 2024-11-25 | $23.45 | Staples | Office Supplies | Operations | Yes (100%) | | **Summary:** | Category | Total | % of Spend | |-----------------------|-----------|------------| | Equipment | $4,200.00 | 61.5% | | Travel & Lodging | $1,682.17 | 24.6% | | Training & Education | $350.00 | 5.1% | | Software & SaaS | $338.99 | 5.0% | | Advertising | $156.00 | 2.3% | | Meals & Entertainment | $67.32 | 1.0% | | Office Supplies | $23.45 | 0.3% | | **Total** | **$6,817.93** | **100%** | ### Example 2: Flagging Misclassified Expenses **Input:** Review existing categorizations for accuracy. **Output — Correction Report:** | Date | Amount | Merchant | Current Category | Corrected Category | Reason | |------------|---------|----------------|------------------|------------------------|---------------------------------------------------------| | 2024-11-07 | $320.00 | WeWork | Office Supplies | Rent & Facilities | Co-working space is r
Design reproducible evaluations for AI agents with representative task sets, explicit rubrics, appropriate graders, baselines, regression gates, and failure analysis. Use when defining agent quality, comparing prompts or models, validating a release, measuring tool-use reliability, investigating regressions, or deciding whether an agent is ready for production.
Design privacy-aware observability for AI agents using traces, spans, structured events, metrics, cost attribution, dashboards, alerts, and investigation workflows. Use when instrumenting an agent, debugging intermittent tool or model failures, defining service-level objectives, analyzing latency or spend, auditing agent decisions, or preparing production monitoring.
Design and verify auditable human oversight, approval gates, escalation paths, and safe state transitions for AI agent workflows. Use when deciding which agent actions require review, adding approve/reject or dual-control flows, preventing unauthorized autonomous effects, creating decision records, reducing rubber-stamping, or recovering safely from rejected, expired, or failed actions.
Design, implement, harden, and verify Model Context Protocol (MCP) servers with precise tool contracts, least-privilege authorization, safe transports, structured errors, and interoperability tests. Use when creating a new MCP server, exposing an API or data source through MCP, reviewing an MCP server design, adding or revising MCP tools, or preparing an MCP server for production.
Design and operate bounded multi-agent workflows with task decomposition, dependency graphs, ownership, handoff contracts, shared-state controls, approvals, recovery, and synthesis. Use when a task contains genuinely independent workstreams, specialized roles, parallel research or implementation, reviewer-worker loops, or coordination problems that one agent should not execute sequentially.
Design and validate model-facing tool definitions with clear names, action-oriented descriptions, bounded JSON Schema parameters, explicit side effects, safe defaults, idempotency, errors, and realistic tests. Use when creating function-calling tools, MCP tools, agent actions, structured tool inputs, or when a model selects the wrong tool, invents arguments, or causes unsafe side effects.
Plan, execute, document, and retest authorized security assessments of AI agents and multi-agent workflows using safe adversarial cases, synthetic identities, canaries, and evidence-based findings. Use when defining red-team rules of engagement, assessing prompt injection or excessive agency, testing tool and identity boundaries, evaluating memory or cross-agent attacks, scoring a campaign, or verifying remediation in an approved environment.
Threat-model and harden AI agents, RAG systems, assistants, and tool-using workflows against direct, indirect, stored, cross-agent, and multimodal prompt injection. Use when reviewing an agent architecture, isolating untrusted content, constraining tools and egress, protecting secrets, adding injection-focused tests, investigating a suspected injection incident, or documenting residual prompt-injection risk.