Trust infrastructure for autonomous agents.
Memory is the substrate. Governance is the product. Novyx Core lets agents remember, coordinate, act safely, and recover — three lines of code.
Get started in 30 seconds
uvx novyx-mcpRead the quickstart →Works locally with zero config. No API key needed.
Four things it does
Remember
Persistent memory with semantic search. Survives sessions, restarts, crashes.
- ✓Sub-100ms recall
- ✓Semantic search
- ✓Knowledge graph
Coordinate
Context Spaces let agents share memory. Agent A learns it, Agent B knows it.
- ✓Shared memory spaces
- ✓91 MCP tools
- ✓Works in Claude Code & Cursor
Act Safely
Policy engine, approval flows, and cryptographic audit for every action.
- ✓Approve before execution
- ✓SHA-256 audit trail
- ✓Risk tiers & delegation
Recover
Point-in-time rollback. Preview before executing. Like git for agent state.
- ✓Restore to any point in time
- ✓Replay time-travel
- ✓Eval health scoring
Get started in 30 seconds
Setup: Install & initialize
pip3 install novyx # or: pip3 install novyx[async]
from novyx import Novyx
nx = Novyx(api_key="YOUR_KEY")Then just 3 lines: remember, recall, rollback
nx.remember("user prefers dark mode")
nx.recall("user preferences")
nx.rollback("2 hours ago")Novyx vs. the alternatives
| Feature | Novyx | LangChain Memory | Vector DB |
|---|---|---|---|
| Persistent across sessions | ✓ | Manual | Manual |
| Point-in-time rollback | ✓ | — | — |
| Cryptographic audit trail | ✓ | — | — |
| Semantic search | ✓ | ✓ | ✓ |
| Knowledge graph | ✓ | — | — |
| Multi-agent sharing | ✓ | — | — |
| MCP server | 91 tools | — | — |
| Threat intelligence | ✓ | — | — |
| Auto defense | ✓ | — | — |
| Circuit breaker | Free | — | — |
| Eval & CI/CD gate | ✓ | — | — |
See it in action
Store memories, corrupt them, roll back to a clean state, and verify the cryptographic proof — all in your browser. No signup required.
Try Interactive Demo →Beyond the basics
Install the MCP server. Add a line to CLAUDE.md. Done.
Every Claude Code session picks up where the last one left off. Your agents share memory, coordinate across sessions, and never repeat the same mistake. Agents act safely — every risky action waits for your approval before it executes.
claude mcp add novyx-memory -- python -m novyx_mcp# CLAUDE.md
## Shared Memory (Novyx MCP)
You have access to novyx-mcp tools for shared memory,
knowledge graph, audit, rollback, and replay.
Store decisions at the end of tasks.
Check for context from other agents before starting.Free
$0/mo
- 5,000 memories
- 5,000 API calls/mo
- 10 rollbacks/mo
- 7-day audit
Starter
$12/mo
- 25,000 memories
- 25,000 API calls/mo
- 50 rollbacks/mo
- 14-day audit
Pro
Most Popular$39/mo
- Unlimited memories
- 100,000 API calls/mo
- Replay, Cortex, 91 MCP tools
- Unlimited rollbacks
Enterprise
$199/mo
- Unlimited everything
- 90-day audit retention
- Priority support
- SSO-ready & 99.9% SLA
No lock-in. Upgrade instantly. Cancel anytime.
What's next
Actively building. Shipped weekly.
Context Spaces
LiveShared memory namespaces for multi-agent teams. Scoped access, cross-agent recall.
Cortex v2
In progressSmarter autonomous maintenance — auto-merge duplicates, decay noise, surface patterns.
Replay UI
Coming soonVisual time-travel dashboard. See exactly what your agent knew at any point. API already live.
SOC 2 Type II
PlannedFormal compliance certification. Building internal controls now.
Self-hosted option
PlannedRun Novyx Core on your own infrastructure for full data sovereignty.
Your AI finally remembers.
Free tier. No credit card. Start in 30 seconds.
5,000 memories free. Forever.
Join our DiscordBuilt on Novyx Core
Teams are building real products on the trust layer. Here's what's running in production.
Your product
Novyx Core is the trust layer for agent infrastructure. pip install novyx and start building.