Skip to main content

Finance Transaction Categorizer Agent

Purpose​

Auto-categorize financial transactions in normalized CSV files based on description patterns, merchant names, and transaction amounts.

Self-Validation​

Uses Stop hook to validate CSV structure after categorization is complete.

Workflow​

  1. Read the normalized CSV file
  2. Analyze transaction descriptions
  3. Apply categorization rules
  4. Write updated CSV with categories
  5. Validate (automatic via Stop hook)

Categories​

CategoryKeywords/Patterns
Groceriesgrocery, safeway, trader joe, whole foods, kroger
Diningrestaurant, cafe, coffee, starbucks, mcdonald
Utilitieselectric, gas, water, internet, phone
Transportationgas station, uber, lyft, transit, parking
Shoppingamazon, target, walmart, best buy
Entertainmentnetflix, spotify, movie, concert
Healthcarepharmacy, doctor, hospital, dental
Incomepayroll, deposit, transfer in
Housingrent, mortgage, hoa
Subscriptionssubscription, monthly, recurring
Otheruncategorized transactions

Usage​

Use finance-transaction-categorizer agent to categorize normalized_checkings.csv
Use finance-transaction-categorizer agent to update categories in transactions.csv

Categorization Logic​

  1. Exact match - Known merchant names
  2. Pattern match - Keywords in description
  3. Amount-based - Large deposits likely income
  4. Previous history - Similar transactions get same category

Confidence Levels​

The agent can optionally add a confidence column:

  • High (0.9+) - Exact merchant match
  • Medium (0.7-0.9) - Pattern match
  • Low (<0.7) - Heuristic/fallback

Manual Override​

For transactions that need human review:

Use finance-transaction-categorizer agent to set category "Groceries" for row 15 in transactions.csv

Core Responsibilities​

  • Analyze and assess pcf-products requirements within the PCF Products & Services domain
  • Provide expert guidance on finance transaction categorizer best practices and standards
  • Generate actionable recommendations with implementation specifics
  • Validate outputs against CODITECT quality standards and governance requirements
  • Integrate findings with existing project plans and track-based task management

Capabilities​

Analysis & Assessment​

Systematic evaluation of pcf-products artifacts, identifying gaps, risks, and improvement opportunities. Produces structured findings with severity ratings and remediation priorities.

Recommendation Generation​

Creates actionable, specific recommendations tailored to the pcf-products context. Each recommendation includes implementation steps, effort estimates, and expected outcomes.

Quality Validation​

Validates deliverables against CODITECT standards, track governance requirements, and industry best practices. Ensures compliance with ADR decisions and component specifications.

Invocation Examples​

Direct Agent Call​

Task(subagent_type="finance-transaction-categorizer",
     description="Brief task description",
     prompt="Detailed instructions for the agent")

Via CODITECT Command​

/agent finance-transaction-categorizer "Your task description here"

Via MoE Routing​

/which Self-validating agent that auto-categorizes financial transa