A collaborative knowledge base that is privacy-focused & AI-native.
Superprism refracts your scattered knowledge into focused, powerful contexts. Organize your information into distinct spaces, ready to work with AI and external tools within your existing workflows.
Privacy-first architecture protects your data. Context isolation keeps your AI sharp.
Join the Waitlist to get early access.
How it Works
Build Spaces of Knowledge
Create dedicated spaces for different contexts.
Upload documents, connect data sources, or ingest content from the web. Superprism organizes it automatically based on rules you set.
Your knowledge stays clean and up to date. No context bleed between unrelated work.
Run Projects Within Spaces
Launch projects that use your space's knowledge as foundation.
Chat with your knowledge base, generate reports, create summaries, build documentation. Projects can pull in additional context without polluting your core space—they're experimental branches where only what you choose to keep persists.
Generate what you need. Keep what matters.
Share and Extend
Invite collaborators and connect to external tools and workflows.
Generate your own Model Context Protocol (MCP) server to connect your organized knowledge to other AI-ready tools. Export selective knowledge as wikis, summarized docs or public documentation. Your knowledge base becomes infrastructure for your entire workflow.
Private by default. Collaborative when you choose. Connect to your existing tools.
Key Capabilities
Dynamic Knowledge Ingestion
Pull knowledge from multiple sources: upload files directly, scrape web content, connect APIs, or sync from external stores like Google Drive or Obsidian vaults. Superprism structures incoming knowledge based on your organizational rules—tags, categories, hierarchies—as it arrives.
Intelligent Knowledge Navigation
Explore your knowledge base through traditional file browsing, semantic search, or visual knowledge graphs. Chat directly with your space to surface relevant context. Edit and refine knowledge blocks, and watch updates propagate through your entire base.
Context-Isolated Projects
Projects give you a sandbox to work with your knowledge base without contamination. Add project-specific context that stays ephemeral unless you explicitly choose to persist it back to your space. Perfect for experimental explorations, client work, or one-off analyses.
Multi-Modal Interactivity
Generate artifacts in multiple formats: documents, reports, summaries, structured data, or custom formats. Engage with outputs through chat interfaces, insight feeds, or visual representations. Export to other tools or publish selectively for external audiences.
Extensible Architecture
Built on open-source foundations. Create custom MCP servers from your spaces to enhance Claude, GPT, or other AI tools with your organized context. Build your own workflows, automations, and integrations. Fork and adapt the system to your exact needs.
Flexible Collaboration
Keep spaces private for personal knowledge gardening, or invite collaborators for team projects. Share selective portions of your knowledge as wikis, documentation sites, or structured exports for partners and stakeholders. You control the boundaries.
Design Principles
Local-First Intelligence
Your knowledge lives on your infrastructure—private by default, sovereign by design. Share selectively when it serves your purpose. Privacy and control aren't features; they're the foundation.
Context Hygiene
Separate spaces prevent unrelated knowledge from contaminating each other. Isolated projects let you experiment without polluting your core knowledge base. Relevance stays high because context stays focused.
Open and Portable
Built on open-source foundations, Superprism is forkable, extensible, and yours to adapt. No vendor lock-in. No proprietary formats. Your knowledge remains portable across tools and workflows.
Symbiotic Productivity
AI handles scale and synthesis. Humans provide judgment, creativity, and direction. Superprism weaves both into a partnership where machines augment rather than replace human intelligence. The system scales collective knowledge without sacrificing individual ingenuity.
Technical Foundation
Transparent, Self-Hostable Stack
Built on Next.js, Supabase (Postgres/Auth/Storage), and our Model Context Protocol server. Deploy on your own infrastructure with complete visibility into where your data lives. Every component is transparent and under your control.
Non-Custodial by Design
Supabase row-level policies keep workspaces isolated. N8n ingestion pipelines register every job in the open schema. You maintain custody of your knowledge base—no data handover required.
Secure Context Streaming
The MCP layer exposes context to AI agents via per-workspace API keys or Supabase OAuth. Stream read-only context to tools without exposing raw data, maintaining security while enabling powerful integrations.
Future-Ready Coordination
Aligning with emerging standards like ERC-8004 (autonomous worlds state sync), AP2's agent-to-protocol graph, and X402 wallet channels. Your spaces can plug into onchain workflows as these coordination rails mature—no custom glue code needed.
Start refracting knowledge soon.
Whether you're a researcher organizing notes and reference material, a product team maintaining context across sprints, or a DAO coordinating community knowledge—Superprism gives you the foundation to work smarter with AI.
Join the Waitlist to get early access.