About Localecheck
Localecheck provides verified locale utilities for AI agents and developers working with UK and US data. It handles dates, phone numbers, currency, tax rates by date, holidays, and addresses, flagging invalid inputs rather than guessing.
Key features
- 01Parse locale-sensitive dates
- 02Validate phone numbers
- 03Format currencies
- 04Look up VAT and sales tax by date
- 05Check public holidays
- 06Parse UK and US addresses
Use cases
- 01Determine historical VAT for an invoice
- 02Format amounts for a locale
- 03Check holidays when planning delivery dates
- 04Validate phone numbers and parse addresses
MCP support
The TypeScript and Python packages provide MCP servers exposing locale parsing, validation, formatting, tax, holiday, and address tools.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
Tools
7Saved tool list · updated 3w ago
| Parameter | Type | Required | Description |
|---|---|---|---|
| input | string | Yes | |
| locale | string | No |
Details
From the project READMElocalecheck
Locale-correctness tools for AI agents. Agents are systematically wrong about
everyday non-US data — they read 03/04 as March 4 in the UK, apply today's VAT
rate to a historical invoice, format €1,234.50 for a German reader, and miss
bank holidays in delivery dates. localecheck gives them verified answers and
flags impossible inputs instead of guessing.
Coverage: UK + USA (v1), expanding. Free, no API key.
Packages
| Package | Install | Use | |
|---|---|---|---|
| TypeScript (primary) | packages/localecheck-ts | npx -y localecheck (live on npm) | MCP server and importable library |
| Python | packages/localecheck-python | from source (PyPI release coming soon) | MCP server, importable library, REST API |
Both wrap the same authoritative libraries (libphonenumber, CLDR/Intl, date-holidays) and share the same hand-curated tax data.
What it does
parse_date · validate_phone · format_currency · tax_rate (VAT/sales-tax
by date) · is_holiday / next_holiday · parse_address
Quick start (in an AI app)
{ "mcpServers": { "localecheck": { "command": "npx", "args": ["-y", "localecheck"] } } }
Then ask naturally — the agent picks the tool: "A UK invoice dated 1 June 2009 — what VAT rate?" → 15%.
Repo map
packages/localecheck-ts— primary TypeScript package (MCP + library)packages/localecheck-python— Python package + FastAPI REST APIeval/— benchmark harness: how often raw LLMs get this wrongbenchmark.html— shareable results page ·BUSINESS_MODEL.md— strategy ·PUBLISH.md— release steps ·LOCAL_TEST.md— test before publishing
Status & caveats
UK VAT history is reliable for the modern era (standard rate). US state sales-tax values are 2025 state base rates (local taxes add on top) — verify against an authoritative feed before live invoicing. Addresses parse + extract a valid postcode/ZIP but don't verify deliverability.
Code: Apache-2.0. Curated tax data: see BUSINESS_MODEL.md.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.