product
The governance layer your AI agents don't have
Agents discover and call tools. Neurelay adds what's missing — identity, policy, enforcement, and audit. The same pattern API gateways brought to REST.
How it fits
Your agents connect to Neurelay instead of directly to tool servers. Neurelay connects to the servers on their behalf — scoped by policy.
Architecture diagram — coming soon
Agent → Neurelay Gateway → Policy Check → Tool Servers
What you control
blueprints
Tool server registry
Register your tool servers as blueprints. Each blueprint captures the server's tools, their input schemas, connection configuration, and metadata.
- · Auto-discover tools from connected servers
- · Version your tool catalog
- · Group related tools by domain
Screenshot — blueprint registry
Screenshot — policy editor
policies
Fine-grained access control
Policies define which tools a credential can access and what operations are allowed. Read, write, execute, or deny — per tool, per credential.
- · Default deny — nothing accessible until granted
- · Reusable policy templates per domain
- · Instant policy updates — no agent restart
gateway
Aggregation & enforcement
One endpoint for your agents. The gateway aggregates tools from multiple servers, filters discovery by policy, and gates every execution.
- · Agents see one endpoint, gateway handles routing
- · Connection credentials stay in the gateway — not in agents
- · Defence in depth — discovery filter + execution gate
Screenshot — gateway dashboard
Screenshot — audit log
audit
Full trail & kill switch
Every tool call logged with who, what, when, and the result. Revoke a credential or disable a tool from the dashboard — takes effect immediately.
- · Compliance-ready audit exports
- · Instant revocation — no agent reconfiguration
- · Incident response with full call history
discovery
Smarter tool discovery
Every tool in the discovery response costs tokens and increases the chance the agent picks the wrong one. We call this the hallucination tax — more tools, worse decisions. Neurelay cuts it at three levels.
- · Policy filtering — agents only see authorized tools, not the full catalog
- · Description tuning — edit tool descriptions from the dashboard, no redeploy
- · Fewer tools in context → fewer tokens → better selection → fewer failed calls
Screenshot — tool description editor
Screenshot — environment connections
environments
Stage & production connections
Connect tool servers per environment. Test policies and agent behaviour in staging before promoting to production — at the governance infrastructure level, not inside your dev framework.
- · Separate connections for staging and production
- · Test policy changes before they affect live agents
- · Promote configurations across environments