Explainer
What MCP is, and why your AI needs it
The Model Context Protocol is how AI tools reach knowledge that isn't in their training data. Here is what it actually does, in plain terms.
Verde
Stop explaining your company to your AI.
Verde is one shared memory for your team's decisions, policies and know-how — and Claude, Cursor, ChatGPT and Codex all read from it. Free forever.
- Working in minutes
- No AI API keys
- Free forever, no card
Speaks the open standard your tools already use
Why it matters
Right now, your best answers are stuck in someone’s chat history.
Your pricing decision lives in one person’s Claude history. Your architecture call is in someone’s Cursor session. Your support process is in a doc nobody updates. So you answer the same question twice, your tools confidently repeat last quarter’s prices, and nobody can say which version of a decision is the real one.
Verde gives all of it one home that your people and your AI tools both read from — and that stays current when things change.
You stop repeating yourself.
Write a decision down once. Every tool you connect reads it from then on.
Everyone gets the same answer.
One shared workspace, one current version, whoever is asking.
Your AI stops being confidently wrong.
Superseded knowledge stays readable but never wins retrieval.
What you get
Three things that make it trustworthy.
One endpoint, every tool
Paste a single MCP endpoint into Claude, Claude Code, Cursor, Codex, or ChatGPT. No plugins, no per-tool setup, and no lock-in — switch AI vendors and your memory comes with you.
Your AI stops quoting old prices
Supersede a decision and the newest published version always wins retrieval. The history stays readable and clearly marked, so you get institutional memory and one current answer.
A private note stays private
Public, company, and private scopes are enforced server-side in SQL on every query, never trusted to the AI client. No owner or admin can browse another member's private memories.
Pricing
Free forever, and not a trial.
No card, no clock, no expiring workspace. Paid plans exist for when your team outgrows one vault — everything below is what $0 gets you, permanently.
- Unlimited documents
- 3 write seats, 5 free readers
- 2,000 AI calls a month
- Every MCP client
- A public KB, up to 10 pages
- No AI API keys required
From the blog
Notes on working with AI memory.
Practice
Every empty search is a document you haven't written
When your AI searches your knowledge base and finds nothing, that's not a miss. It's a question your team never wrote down — and the most useful backlog you have.
Verde
Engineering
Who can actually read your private notes
Private, company, public — three scopes, one enforcement point. Why the check has to live in the query, and not in the AI client's good behaviour.
Verde
Stop being your AI’s memory.
Create a workspace, invite your team, and connect your first AI tool before your coffee gets cold.