Protect AI agent memory with zero-knowledge encryption and differential privacy. Built for developers who need secure, privacy-preserving context storage across multi-turn conversations without exposing sensitive data to third parties.
io.github.SynapseLayer/synapse-secure-memory
Local install
STDIO
No auth required
How models use it and what it is built for.
Protect AI agent memory with zero-knowledge encryption and differential privacy. Built for developers who need secure, privacy-preserving context storage across multi-turn conversations without exposing sensitive data to third parties.
Local install — runs as a subprocess.
Configuration this server reads at startup.
Identifier for the agent using the memory layer (e.g., claude-desktop, my-agent).
Differential privacy epsilon parameter. Lower values = stronger privacy. Range: 0.1–2.0.
AES-256-GCM encryption key for at-rest memory encryption. Auto-generated if not provided.
Logging verbosity: DEBUG, INFO, WARNING, ERROR. Default: INFO.
Where to find authoritative docs and source for Synapse Layer — Zero-Knowledge Memory.
Paste any of these into Agent Studio after connecting Synapse Layer — Zero-Knowledge Memory.
Common questions about connecting and running Synapse Layer — Zero-Knowledge Memory.
What does Zero-Knowledge Context mean in this MCP server?
Zero-Knowledge Context™ means the memory layer can verify and use encrypted agent context without decrypting or exposing the underlying data. Your sensitive conversation history stays encrypted at rest and in transit.
How do I set the privacy level for my agent?
Use the SYNAPSE_PRIVACY_EPSILON environment variable (range 0.1–2.0). Lower values like 0.1 provide stronger differential privacy guarantees; higher values like 2.0 allow more utility but less privacy. Start with 0.5 for balanced protection.
Do I need to provide my own encryption key?
No. If you don't set SYNAPSE_ENCRYPTION_KEY, the server auto-generates an AES-256-GCM key for you. You can supply your own key for key management integration with your infrastructure.
What is Neural Handover and when would I use it?
Neural Handover™ is the mechanism for securely transferring context between AI agents without exposing raw memory. Use it when routing tasks between specialized agents or when handing off conversation state while maintaining privacy guarantees.
How do I troubleshoot memory encryption or privacy issues?
Set LOG_LEVEL=DEBUG to see detailed encryption and differential privacy operations. Check that SYNAPSE_AGENT_ID is unique per agent and SYNAPSE_ENCRYPTION_KEY is consistent across restarts to avoid decryption failures.
MCP Playground runs 10,000+ hosted MCP servers — GitHub, Linear, Notion, Stripe, Sentry and more — across Claude, GPT, Gemini, DeepSeek and 60+ AI models. Compare model answers side-by-side, save agent presets, share runs. Zero install.
Open Agent StudioJudges Panel
45 judges that evaluate AI-generated code for security, cost, and quality with built-in AST.
Addozhang Nexus
Query Sonatype Nexus Repository (OSS/Pro) for Maven, Python, and Docker artifacts
Cyanheads MCP Ts Template
TypeScript template for building MCP servers with declarative tooling, observability, and auth.
Chernistry Bernstein
Declarative agent orchestration for engineering teams