ezbookkeeping_skill

Overview

This skill integrates with ezBookkeeping via the ezBookkeeping API Tools to let AI agents manage personal finance data on a self-hosted instance. It exposes commands for adding and querying transactions, accounts, categories, and tags so agents can automate bookkeeping tasks. The skill is lightweight and relies on two environment variables for server address and API token.

How this skill works

The skill invokes the ezBookkeeping API Tools script on the host (sh or PowerShell) and forwards commands to the configured ezBookkeeping server. It inspects and manipulates core bookkeeping entities—transactions, accounts, categories, and tags—by calling the tool commands and parsing the responses. Authentication and endpoint addressing are handled by environment variables so the agent does not need to manage low-level HTTP details.

When to use it

• Automate recurring transaction creation and reconciliation.
• Build conversational workflows that query balances, recent transactions, or category spending.
• Sync or export small personal finance datasets to other systems.
• Add tags or categories programmatically for bulk categorization.
• Query account or category summaries for reporting or budgeting.

Best practices

• Set EBKTOOL_SERVER_BASEURL and EBKTOOL_TOKEN securely as environment variables before use.
• Test commands in a sandbox or with a small dataset to confirm expected behavior.
• Use descriptive categories and tags to make query results more actionable.
• Limit write operations in automated flows; prefer review steps for mass changes.
• Log API tool outputs and errors to help troubleshoot failed operations.

Example use cases

• An agent adds a week of transactions from a receipt OCR pipeline into ezBookkeeping automatically.
• A chatbot answers “How much did I spend on groceries last month?” by querying category totals and returning a short summary.
• Batch-tagging older transactions based on merchant rules produced by an AI rule engine.
• Export account balances to a budgeting assistant every morning to update forecasts.
• Create a new account and seed opening balance when a user connects a new bank in a migration script.