companieswise
Provides verified UK company lookup and number validation for AI agents using official Companies House data. Enables lookup of registered details by number, validation of company number format, and search by company name.
README
<div align="center">
<img src="https://qinisolabs.github.io/companieswise/logo.svg" width="96" height="96" alt="Qiniso" />
companieswise
Verified UK company lookup & number validation for AI agents — official Companies House data, not guesses.
Verified, trustworthy data tools for AI agents. "Qiniso" means "truth" in Zulu.
Website · npm · MCP Registry
</div>
Ask an LLM "what's the registered name and status of company 00445790?" and it will answer confidently — and usually wrongly: invented names, the wrong status, a fabricated incorporation date. Company records live in the Companies House register, not a model's weights. companieswise looks the number up in the official Companies House snapshot and returns the real registered details — or an honest "not found" instead of a guess.
⚠️ Read this first — what the data is
- UK-wide (England & Wales, Scotland, Northern Ireland).
- A monthly snapshot of the live register. Company status is as of the dataset date, not real-time — a company dissolved last week may still read "Active" until the next snapshot, and dissolved companies are generally absent. For current status, check the live register.
- Not advice. Not legal, financial, credit, or KYC/AML advice, and not a substitute for statutory due diligence.
Install
{ "mcpServers": { "companieswise": { "command": "npx", "args": ["-y", "companieswise"] } } }
Then load the real data once (the package ships with only a small sample):
npx -p companieswise companieswise-update
This downloads the latest monthly Companies House snapshot (~tens of MB) to a local cache. Re-run it whenever you want to refresh; a GitHub Action rebuilds the snapshot monthly, so companieswise-update always fetches the current month. Until you run it, validate_company_number works fully and lookup_company/search_company clearly say they're on the sample.
Optional: live mode (BYO key)
By default companieswise serves the monthly snapshot (offline, no key). If you set a free Companies House API key, it switches to live queries against the official Companies House API instead — real-time register status, including dissolved companies, on your own key (no shared rate limit):
{
"mcpServers": {
"companieswise": {
"command": "npx",
"args": ["-y", "companieswise"],
"env": { "COMPANIESWISE_CH_API_KEY": "your-companies-house-api-key" }
}
}
}
lookup_company and search_company use the live API when the key is set (results tagged "dataset": "live") and fall back to the snapshot when it isn't ("dataset": "snapshot"), so it's always clear which you're getting. validate_company_number is offline either way. The key is sent only to the official Companies House API. CH_API_KEY is accepted as an alias.
Use it as a library
npm i companieswise
import { lookupCompany, validateCompanyNumber, searchCompany } from "companieswise";
validateCompanyNumber("SC123456"); // { valid: true, prefix: "SC", meaning: "Company registered in Scotland" }
validateCompanyNumber("6").number; // "00000006" — normalises/zero-pads
lookupCompany("00000006"); // official name, status, type, incorporation date, postcode, SIC — or found:false
searchCompany("greggs"); // companies whose registered name contains your words
A well-formed number that isn't in the snapshot returns found: false with a clear note — it never invents a company.
Tools — 3
| Tool | What it answers |
|---|---|
| lookup_company | A company's official registered details by number (name, status, type, incorporation date, registered postcode, primary SIC) |
| validate_company_number | Is this a well-formed UK company number, and what does its prefix denote? (format only — there is no check digit) |
| search_company | Reverse lookup — find a company's number from its name |
Data & monthly auto-refresh
The data is the Companies House Free Company Data Product (UK-wide, live register), published under the Open Government Licence v3.0. Because the real dataset is ~5.6M companies — far too large to bundle into an npm package — the design splits code from data:
- Code ships on npm (tiny).
- Data is built monthly by a GitHub Action (
.github/workflows/refresh-data.yml) that downloads the official snapshot, compiles a compact artifact, and publishes it as a GitHub Release asset. companieswise-updatedownloads that artifact into a local cache; every response reports itsdatasetVersion(the snapshot date).
No API key, no rate limit, no per-request cost — lookups run locally against the cached snapshot. See ARCHITECTURE.md for why, and the hosted/real-time path.
What it is not
- Not real-time. Monthly snapshot of the live register; status is as of the dataset date. For live status use the Companies House API/website.
- Not the full register. The free snapshot is live companies; dissolved companies are generally absent (absent ≠ "never existed").
- Not advice, and not a statutory KYC/AML check.
- Not a guesser — unknown numbers return an honest "not found".
License
Apache-2.0. Company data © Crown copyright, Companies House, Open Government Licence v3.0; see NOTICE.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.