Alion MCP — company, startup and market intelligence
Alion is a live graph of companies worldwide, built from the companies' own websites, job boards and public registers. This MCP server gives Claude, ChatGPT, Cursor and any MCP client read-only access to it:
- what a company is — what it does, where it operates, its size, founding year and offices;
- how it is funded — funding rounds, amounts, dates, investors and accelerators;
- what it builds with — the technologies named in its own job postings;
- how fast it grows — hiring momentum, new offices and technologies, a dated story of its stages;
- who is like it — lookalike companies with the reasons they are similar;
- the labour market — verified job postings and pay, demand and time-to-fill statistics.
Use it for market maps, competitor and lookalike lists, sales prospecting, startup and deal-flow scouting, due diligence on a company, technology-adoption research and labour-market questions.
Server: https://alion.io/mcp (Streamable HTTP, OAuth 2.1) · Docs: https://alion.io/developers/agents
This repository holds the public manifests of the hosted server (MCP Registry
server.json, an Agent Pluginsplugin.json+mcp.json) and the connection guide. The server itself runs at alion.io; there is nothing to install or self-host.
Tools
All tools are read-only (readOnlyHint: true): nothing is created, changed or sent.
| Tool | What it answers |
|---|---|
search_companies |
Companies by industry, country, size, founding year, startup, funding stage, rounds raised since a date, minimum round, investor or accelerator, technology, growth level |
similar_companies |
Lookalikes of a company: shared industry segments and technologies, same headquarters country, a near size band |
get_company |
One company in depth: offices, funding rounds and investors, backers, stack, hiring and growth, traction, a dated story of its stages |
resolve_company |
The company behind a name, a domain or a URL, and whether it is alive |
search_jobs |
Live job postings verified on the employers' own boards, with pay (stated or estimated) |
get_job |
One posting in full: requirements, technologies, benefits, lifecycle |
market_stats |
Pay percentiles, demand and its trend, time to fill and top employers for a slice of the job market |
The server also exposes 5 prompts (company_research, market_map, startup_scout, market_brief, job_shortlist) and 5 dictionary resources (alion://dictionaries/...: industries, countries, role families, ATS vendors, fixed vocabularies).
Try
- "Find companies similar to mambu.com and explain what they have in common."
- "Which early-stage startups in Germany raised a round in the last year?"
- "Tell me about finmid.com: funding and investors, tech stack, and how fast it is growing."
- "Fintech companies in the UAE with 51-200 people that use Kubernetes."
- "What do backend engineers earn in Germany, and which companies hire the most of them?"
Every answer links to the Alion page of each company or job posting.
Connect
Sign-in is OAuth with your Alion account (free); the client opens the sign-in page on first use.
| Client | How |
|---|---|
| Claude (claude.ai, Desktop) | Settings → Connectors → Add custom connector → URL https://alion.io/mcp → Connect |
| ChatGPT | Settings → Apps & Connectors → Advanced → Developer mode → Create → URL https://alion.io/mcp, authentication OAuth |
| Claude Code | claude mcp add --transport http alion https://alion.io/mcp, then /mcp to sign in |
| Cursor | add to ~/.cursor/mcp.json: {"mcpServers": {"alion": {"url": "https://alion.io/mcp"}}} |
| VS Code (Copilot) | Command palette → MCP: Add Server → HTTP → https://alion.io/mcp |
| Cline | MCP Servers → Configure → add "alion": {"type": "streamableHttp", "url": "https://alion.io/mcp", "headers": {"Authorization": "Bearer <key>"}} (step by step: llms-install.md) |
| Codex | codex mcp add alion --url https://alion.io/mcp, then codex mcp login alion |
| Mistral Le Chat | add a custom MCP connector with URL https://alion.io/mcp and OAuth sign-in |
| Any MCP client | Streamable HTTP endpoint https://alion.io/mcp; OAuth discovery at https://alion.io/.well-known/oauth-protected-resource/mcp (dynamic client registration and client ID metadata documents are supported) |
Scripts can use an API key instead of OAuth: Authorization: Bearer <key> from https://alion.io/developers/keys. Every tool is also a plain HTTP endpoint (/mcp/rest/<tool>), described in OpenAPI 3.1 at https://alion.io/mcp/openapi; the live tool catalogue with input and output schemas is at https://alion.io/mcp/tools.
Pricing
- Compact lists (
detail="compact", up to 20 rows) ofsearch_companies,similar_companiesandsearch_jobsare free up to 50 a day for an assistant connected through OAuth. - Full company profiles, job records and market statistics are paid from Alion credits (1 credit = $0.10; a company profile costs a few cents). Every result carries
metawith what the call cost and what is left. resolve_companyis free. Prices: https://alion.io/developers/pricing
Data and privacy
Company and posting texts are third-party content: they are filtered for text aimed at AI models before they are served, and the tools present them as data, never as instructions. Contacts of people are not exposed. The server logs the tool name and its arguments (capped at 512 characters), the client name, status and duration for quotas, billing and abuse prevention, kept 12 months; it never reads conversation content, memory or history.
- Terms and privacy for the MCP server and API: https://alion.io/api-terms
- Privacy policy: https://alion.io/privacy · Terms of service: https://alion.io/terms
Listings
- MCP Registry:
io.alion/alion - Smithery: https://smithery.ai/servers/alion/company-intelligence
Support
https://alion.io/contact · [email protected]
License
The manifests and documentation in this repository are MIT licensed (see LICENSE). The Alion service and its data are covered by https://alion.io/api-terms.
