Works with the AI Agents you already use
Claude Code
Cursor
Codex CLI
Muse
Dots+22 moreClaude CodeCursorCodex CLIMuseDots
OpenClaw
Manus
Grok
Grok Bot+18 agentsZenMoney Receipt Categorization with MCP
Match receipts to ZenMoney expenses, preview category splits, and verify approved changes through your own MCP connector.
$5
/moSubscription · cancel anytime
Secure checkout via Stripe
- Always the latest version
- Delivered live to your agent via MCP
- Cancel anytime, access ends at period end
ZenMoney Receipt Categorization with MCP
Example session with this skill installed
Synthetic example — no personal financial data:
Categorize this example receipt in ZenMoney. It is dated 2026-01-10, paid from my confirmed EUR Demo account, and contains bread €3.00, fruit €5.00 and dish soap €2.00. Total €10.00, no discount. Show the exact preview before writing. Assume the demo connector has no matching expense and already has Food and Household categories.
- Read your context and instructions
- Compiled the zenmoney receipt categorization
Synthetic example: preview before approval
This is an illustrative output using fictional receipt facts and a demo connector, not a recording of a customer's account or proof of a live write.
The example receipt totals €10.00. The demo search has no matching expense. The confirmed account is Demo (EUR), and both proposed categories already exist.
| Date | Account | Category | Receipt items | Expense |
|---|---|---|---|---|
| 2026-01-10 | Demo (EUR) | Food | Bread €3.00; fruit €5.00 | €8.00 |
| 2026-01-10 | Demo (EUR) | Household | Dish soap €2.00 | €2.00 |
| Total | €10.00 |
The proposed action creates two expenses, one per supported category. No expense has been created yet. No new category or transaction comment is proposed.
In this example, receipt memory is disabled, so no local purpose groups will be retained. Enabling it is a separate setting with its own preview and retention choice.
Next step
Confirm this exact account, date and category split to apply it, or ask for a correction. After approval, the connector checks for concurrent changes and verifies the resulting expenses before reporting success. If an apply is interrupted, inspect its durable operation record before retrying.
Subscription · always the latest version
- Secure checkout via Stripe
- Cancel anytime
- Delivered live via MCP (optional)
About this skill
Turn receipts into reviewable ZenMoney entries
Attach a receipt image or PDF and let your agent check existing expenses, identify supported categories, and prepare an exact preview. Review the account, date, currency, amounts and categories before an approved write. Missing expenses are created as one expense per supported category.
What is included
- A focused SKILL.md workflow for matching, categorizing and reconciling receipts.
- Setup instructions for the separately installed open-source zenmoney-receipts MCP server.
- Storage/privacy guidance, synthetic examples and the MIT license.
- Recovery guidance for interrupted operations and verification before reporting success.
Optional evidence memory
Opt-in memory keeps narrow approved purpose groups, counts, subtotals, month and category references. It is off on fresh installations. The agent reports its actual storage location and retention. Local mode stores evidence on the MCP machine; hosted mode uses server tenant storage. Receipts and OCR are not retained by this memory feature. Your agent provider still processes the files you give it under its own policies.
Before you start
Install and configure zenmoney-receipts with Node.js 20.11+, npm, Git and your own ZenMoney account. Your agent must support MCP and reading your image/PDF. This skill does not include an OCR engine, a ZenMoney subscription, or a public hosted connector.
Amounts written to ZenMoney are kept distinct from optional full-price evidence. Pre-discount evidence requires a connector that advertises evidencePricing; older versions must report that limitation instead of dropping items.
Control and limitations
Ambiguous matches and unreadable amounts need clarification. Category changes are scoped and previewed; no generic transaction deletion or unrestricted patch tool is exposed. Matching reduces duplicate risk but cannot guarantee bank sync completeness. Host/model receipt extraction can make mistakes: inspect the preview.
Free and MIT-licensed. Source and setup. Report bugs in the repository with synthetic examples; never post personal receipts or credentials.
How to use
Skills on Agensi run inside your AI agent. Copy the prompt, paste it into your agent, and it connects to Agensi and installs this skill for you. If a payment is needed, your agent shows it to you first.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill passes our trust checks below.
Security scanned
Passed our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free forever
No account required to browse
Trust & safety
Security scanned
Verified clean 2 days ago
Requires
Zenmoney
Git
Node.js
macOS