VentureOS
Provides startup verification tools including domain/package/org availability checks, unit economics, runway, market size, and cap table calculations with transparent formulas and warnings to prevent common agent errors.
README
VentureOS — The Reality Check for Agent-Built Companies
An OKX.AI A2MCP service. An agent will tell you the domain is free. It didn't look. It can't look.
Ask a language model to start a company and it will do a beautiful job of the parts that are writing: the name, the positioning, the pitch, the launch tweet. Then it will hand you a domain it never checked, an LTV it computed wrong, and a cap table where it subtracted the dilutions.
On an agent marketplace, the caller is already a language model. It doesn't need help writing. It needs the two things it cannot do:
- Look something up in a registry. A model has no eyes on RDAP. It will still tell you
yourstartup.comis available, with total confidence, because that sentence is what the pattern demands. - Do arithmetic that compounds. Two 20% rounds do not leave founders with 60%. LTV at 0% churn is not "roughly $4,800". Runway is not cash ÷ burn the moment you plan to hire.
VentureOS does exactly those two things, and nothing else.
Tools
| Tool | What it does |
|---|---|
check_name |
Domains via RDAP (the registry, not a guess), package name via the npm registry, org handle via the GitHub API. |
unit_economics |
Contribution, lifetime, LTV, LTV:CAC, CAC payback — each returned with the formula that produced it. |
runway |
Months, zero-cash date, break-even revenue. Steps burn forward month by month, so growing burn is handled correctly. |
market_size |
Bottom-up TAM/SAM/SOM from a countable number of buyers and a price they'd actually pay. |
cap_table |
Founder ownership through each round — and the wrong answer that subtracting would have given. |
No LLM. No API key. Nothing to configure.
The three things it refuses to do
It will not say "available" when it means "I couldn't check."
If RDAP is unreachable, or a TLD's registry refuses the query, the result is status: "unknown", available: null. Never true. A false available is the expensive direction of this error — it's the one that gets someone to design a logo, print cards and register a company. There is a test that cuts the network and asserts that every single check comes back unknown, that none come back free, and that the verdict says "could NOT be verified" rather than staying silent.
It will not invent an LTV when churn is zero.
Zero churn means customers never leave. The lifetime is infinite, so the lifetime value is infinite. There is no finite number to report, and a model asked for one will produce a large, confident, entirely fictional figure. unit_economics returns ltv: null and explains why.
{
"ltv": null,
"lifetimeMonths": null,
"warnings": [
"Monthly churn is 0%, which implies customers never leave and lifetime value is mathematically infinite. No real business has zero churn — LTV is reported as null rather than as a fabricated finite number."
]
}
It will not let compounding pass as addition.
{
"finalFounderPct": 64,
"naiveSubtractionAnswer": 60,
"formulas": {
"ownership": "ownership × (1 − dilution), applied round by round",
"wrongButCommon": "startingPct − sum(dilutions) ← this is the mistake"
}
}
Both numbers come back. The mistake is shown, not just avoided — so an agent that was about to make it can see what it was about to say.
Every number carries its formula
Not decoration. A founder acting on an unexplained figure has no way to catch the moment an assumption was silently wrong, and neither does the agent relaying it. The formula is the audit trail.
{
"ltv": 1600,
"cacPaybackMonths": 5,
"formulas": {
"ltv": "(arpu × grossMargin) / monthlyChurn",
"cacPaybackMonths": "cac / (arpu × grossMargin)"
},
"warnings": ["CAC payback is 19.2 months. Anything past ~12-18 months means growth must be financed rather than self-funded."]
}
The warnings array is the part a model would smooth over. The MCP instructions tell the calling agent to surface it, every time.
A bug worth admitting
The first version of runway computed 11 months where the answer was 12. It stopped counting the month in which the cash ran out.
Off-by-one is a rounding error in most code. Here it is a founder starting their fundraise a month late, on a number a service told them to trust. Fixed, with a test — and a comment explaining which direction of the error is the fatal one.
Run it
npm test # 47 assertions, offline
npm run test:live # hits RDAP, npm and GitHub for real
npm run dev # http://localhost:8789/mcp
No dependencies. No build step. Node 18+.
Deploy
vercel --prod
Your MCP endpoint is https://<your-domain>/api/mcp.
Call it
curl -X POST https://<your-domain>/api/mcp \
-H 'Content-Type: application/json' \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{
"name":"check_name","arguments":{"name":"your-startup"}}}'
Pricing
Free.
MIT.
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.