vinculum
substrate memory for ai
Product
How it works
The substrate, end to end
The field
Parallel sessions, one memory
Mission Control
Live fleet telemetry
AI failure modes
What we design against
Why it exists
The problem in one page
See the dashboard
Watch a live fleet coordinate through one chat window.
Docs
Get started
Quickstart
Zero to first grunt
Architecture
How it's built
Self-hosting
Your box, your data
API reference
Endpoints & tools
Connect your client
Claude Code
Claude.ai
Cursor
Cline
Zed
The Standard
The specification
Overview
The open memory spec
v0.1 specification
Current conformance spec
Revisions
How the spec evolves
Contributing
Propose a change
The encyclopedia
Substrate memory
The definitional reference
vs. vector memory & RAG
Similarity vs. causality
The cost of keeping the thinking
Cache vs. substrate economics
Why LLMs forget
The intuitive entry point
Cookbook
The library
Browse recipes
Ready-to-run prompts, asks & gates
Verified
Curated & safety-audited
Community
Shared by the field
Author
Standing Orders
Rules every session follows
Safe by design
Four-tier safety gates
Publish a recipe
Share a workflow you trust
Pricing
Hosted plans
Free
$0 · solo demo, self-host free
Substrate
$5/mo · hosted memory + dashboard
Pro
$20/mo · spawn parallel workers
Team
$100/mo · shared projects, SSO
More
Enterprise
On-prem · SAML · SLA
Self-host
Free forever · AGPL v3
Compare plans
Full feature matrix
Resources
Company
Blog
Field notes
Changelog
What shipped
About
Who's building this
Trust
Security
How we protect data
Status
Live uptime
Contact
Talk to us
sign in
☰
home
›
blog
Blog
Nothing here yet.