AI is powerful. But it keeps forgetting you.
Every time you open a new AI chat, you start from zero. Your context, your preferences, your knowledge — gone.
Every AI conversation, you:
Your knowledge is:
The result?
AI that could be brilliant for you keeps giving you generic answers — because it doesn't know you.
AI is only as good as the context you give it
There's a missing layer between what you know and what AI can use.
That's exactly what Viewert is.
Meet Viewert
Two primitives. Infinite leverage.
Vellums
Atomic knowledge units — the structured building blocks of your context. Write your ideas, notes, and workflows once.
Librams
Curated bundles of Vellums — your reusable context packages, ready to load into any AI.
MCP + Export
Connect Librams to any AI via MCP — Claude, Cursor, Windsurf, and any MCP client
Export as Markdown — use anywhere, including Git workflows. No lock-in.
From scattered notes to AI-ready context
Write once. Connect to any AI. Your context loads automatically — every session, every tool.
Write your knowledge
Create Vellums — rich-text docs for anything you keep re-pasting: your coding style, brand voice, research notes, project specs, or study guides.
Bundle into a Libram
Group related Vellums into a Libram — a curated context bundle. Toggle each Vellum's AI switch to control exactly what the AI sees. No noise.
Connect to any AI
Connect via OAuth MCP for Claude & Cursor, or use an API key for any MCP client. Paste your context URL into ChatGPT, Gemini, or Grok. One setup — every tool, every session.
One context infrastructure. Every AI.
Build your knowledge graph once. Deliver it to any AI tool — however they accept context.
Claude Desktop
Add the MCP URL and sign in once. Claude auto-discovers your Librams and loads any on request — no API key needed.
Cursor & Windsurf
Add the MCP URL to your project config. Sign in once via browser — your Librams load in every coding session.
ChatGPT
Paste your Libram's public context URL. ChatGPT reads your knowledge instantly — no account needed for public Librams.
Gemini
Paste your context URL directly or inject Libram markdown as the system prompt in AI Studio.
Grok
Grok's large context window handles full Librams. Paste URL or markdown — your knowledge loads in one message.
Llama / Local Models
Fetch your Libram as markdown with curl and inject as system prompt. Fully private — your notes stay on your machine.
The Knowledge Layer
Between Humans and AI
Turn what you know into something AI can use — and something others can discover and experience.
Vellums
Atomic knowledge documents. Write once in rich text, stored as structured JSON, delivered to any AI as clean Markdown.
Graphs
Connect Vellums into a living knowledge graph. See how your ideas relate, cluster, and compound as your understanding grows.
Librams
Curated Vellum bundles — your reusable context packages that load into any AI in one command via MCP or API key.
Hall Posts
Share insights and spark conversations in the Hall — the community feed where knowledge goes public and builds your reputation as a thought leader.
Videos
Record walkthroughs, demos, and voice notes that bring your Vellums to life — knowledge in motion that others can watch, reference, and learn from.
Obsidian helps you think.
Viewert helps AI think with you.
Same question. Completely different answer.
Pick your role and see exactly what changes when AI finally knows who you are.
AI has no idea what you built, who you're selling to, or what you charge. Back to square one.
AI knows your product, ICP, pricing, and brand voice. Sends you a draft you can actually use.
Give AI your perspective
No complicated config. One command connects Viewert to Claude, Cursor, Windsurf, or any AI tool.
Sign up free
Create your account and build your first Libram. Takes 30 seconds.
Connect your AI
Add the MCP URL in Claude, Cursor, or Windsurf — OAuth connects your account automatically.
Ask AI anything
Tell Claude or Cursor to load your Libram. Your full context is available instantly — no pasting, no re-explaining.
No credit card · Free forever plan · Works with any AI
Built for the Agent Economy
AI agents don't just chat — they act. And before they act, they need to know who they're working for. Your Librams are the briefing documents that make agents precise from the first instruction.
Instead of re-briefing every agent from scratch, your Libram loads once — and every agent that touches your work starts with full context.
Research Agent
Reads your annotated sources and argument structure before synthesizing a single word — no more re-summarizing weeks of reading.
Code Agent
Loads your stack, conventions, and architectural decisions before opening a file — codes in your exact style from line one.
Writing Agent
Absorbs your tone, audience, and vocabulary before drafting — every piece sounds like you, not a generic template.
Support Agent
Knows your product inside-out before the first ticket — answers are specific, accurate, and on-brand without a single training run.
Analysis Agent
Understands your metrics, definitions, and reporting standards before running a single query — no ambiguity, no wrong baselines.
Personal Agent
Knows your schedule, preferences, and working style before managing anything — acts like an assistant who already knows you well.
Trusted by creators from
Real Vellums. Real context.
These are the kinds of knowledge documents people build in Viewert and load into their AI tools every day.
Microeconomics Final Exam Study Guide
Course Context
ECON 101 — Introduction to Microeconomics
Final Exam: December 15th | Format: 40 MC, 3 short answer, 1 essay
Topics I'm Struggling With
- •Price elasticity calculations (especially cross-price)
- •Game theory and Nash equilibrium
- •Welfare economics (deadweight loss)
My Learning Style
- •I learn best through worked examples
- •Analogies help me remember concepts
- •I need to understand the "why" not just the "how"
What I Need
Generate a study guide with concept summaries, worked examples, and practice questions.
Built for people who work with AI every day.
If you use AI to build, write, research, or learn — Viewert is the context layer that makes it work like it already knows you.
Developers
Load your codebase conventions, architecture decisions, and API docs into Cursor or Claude. AI codes in your style from the first line.
Researchers
Bundle annotated paper summaries into a Libram. ChatGPT or Gemini synthesises your entire literature review in minutes.
Writers & Marketers
Store brand voice, audience personas, and tone guidelines. Every AI draft comes out on-brand without repeated instructions.
Students
Build subject Librams for exam prep. Load them into any AI tutor — no more re-explaining your course context.
Product Managers
Keep user stories, acceptance criteria, and specs in a Libram. AI coding sessions start with full product context automatically.
Anyone re-pasting notes
If you've ever typed "here's some context about me/my project" into an AI chat — Viewert automates that forever.
Knowledge in public. Reputation in the making.
Share what you know, vote on what works, and discover the builders who think like you.
See what people are building.
Real Vellums from real builders — browse for inspiration or fork a starting point.
Knowledge Should Be Experienced,
Not Just Stored
The best knowledge doesn't sit in a folder. It gets shared, discovered, experienced — and it builds your reputation in the process.
Private Knowledge
Your Vellums and Librams are private by default. Your research, architecture decisions, and brand voice — all yours, all secure.
Public Experience
When you choose to share — your expertise reaches others. Write about products you've used, vote on your experience, and become the trusted signal others follow.
“I used to copy-paste my codebase conventions into every Cursor session. Now I load a Libram and Cursor just knows. It's like it finally has memory.”
— Backend engineer, Series B startup
“Finally a notes app built for AI-first workflows. The Libram format is exactly what I needed to stop re-explaining my research to Claude every session.”
— PhD researcher, Stanford
“My AI copywriting went from "pretty good" to indistinguishable from my own work. Viewert loads my brand voice, and the AI speaks in my voice immediately.”
— Founder, bootstrapped SaaS
We Built Viewert
for Ourselves First
Every AI session started the same way. Open Claude, open Cursor, and immediately start copy-pasting: the codebase conventions, the project context, the notes from the last sprint. Every. Single. Time.
We were re-explaining our own product to an AI 40 times a day. The AI wasn't getting smarter about us — we were just getting faster at being repetitive.
So we built Viewert. A place to capture knowledge in a format AI can actually use — and reuse. And once we stopped re-explaining ourselves, the quality of every AI session changed completely.
“The competitive advantage wasn't which AI we used. It was the quality of context we gave it.”
We're opening it to everyone who works with AI and knows there's a better way. Join us.
The Future Is Context
AI is getting smarter every year. But the real competitive advantage was never the model — it was always the quality of context you give it.
The builders who win won't just use better AI. They'll know how to brief it perfectly. Every time.
Viewert is the infrastructure for that context layer.
Stop re-explaining
yourself.
Build your knowledge layer once. Every AI you work with loads it automatically — every session, forever.