◆ Igris Sentinel
Govern Every MCP Tool Call Your AI Agents Make — in Real Time
Igris Sentinel is the MCP gateway from Igris Security. Add the SDK to your AI agent application and every tool call is instantly governed: RBAC policies enforced, upstream credentials injected, unauthorized calls blocked, and every interaction logged — with less than 1ms of overhead and a single function call.
SDK-First
Integration
Deny-by-Default
Security
<1ms
Latency
AES-256
Encryption
RBAC
Policies
◆ Capabilities
Built for Runtime Governance
Seven core capabilities that give your security team complete control over what your AI agents can and cannot do
MCP Connections — Encrypted Credential Vault
Store upstream API keys, tokens, and certificates in the Igris Sentinel credential vault — encrypted at rest with AES-256-GCM. Your SDK users and agents never see real upstream secrets.
RBAC Policy Engine
Define who can call what, and under what conditions. First-match-wins rule evaluation with metadata conditions on role, team, environment, and any custom field. One policy, enforced on every call.
Deny-by-Default Security Posture
No rule = no access. Every MCP tool call must be explicitly allowed by a policy rule. Unknown tools, unknown users, and unrecognized conditions are blocked automatically.
Credential Injection
Igris Sentinel strips your users' API key from outbound requests and injects the secure upstream credential from the vault. Your MCP users only ever interact with Sentinel — never with the upstream server directly.
Real-Time Call Blocking & Session Kill Switch
Block individual tool calls based on policy, or instantly suspend an entire user session from the Igris Lens dashboard. Stop suspicious activity in seconds without touching your agent code.
Token Usage & User Management
Track token consumption per user, per connection, and per team. Set budget limits. Manage user access, roles, and onboarding directly from the Igris Security dashboard.
Complete Audit Trail
Every tool call is logged with user identity, trace ID, tool name, action taken, policy decision, and latency. Filter, search, and export directly from Igris Lens.
Alert Dispatch
Real-time alerts via Slack and Discord webhooks. Configurable per event type: policy denials, anomalies, session suspensions. Block Kit formatting for Slack, embed formatting for Discord.
Anomaly Detection
Detects unusual tool call patterns using rolling window statistics. Volume spikes, unusual tool usage, and pattern deviation — all flagged automatically without blocking the request path.
◆ How It Works
Running in 3 Steps
From SDK install to governed MCP calls in under five minutes
Install the Igris Sentinel SDK
One command adds Igris Sentinel governance to your application. The SDK handles authentication, config generation, connection management, and automatic retries.
# Install the SDK
npm install @igris/sdk
# Initialize
import { Igris } from "@igris/sdk";
const igris = new Igris({
apiKey: process.env.IGRIS_API_KEY,
});Define Your RBAC Policies
Create deny-by-default policies with metadata conditions using the Igris Security JSON policy DSL or the no-code policy builder in the dashboard. Policies take effect immediately — no restart required. Or configure policies visually in the Igris dashboard — no code required.
// Policy with metadata conditions
{
"name": "Intern safety guardrails",
"connection": "vk_github_prod",
"rules": [
{
"tool": "delete_*",
"action": "deny",
"conditions": { "role": "intern" }
},
{ "tool": "*", "action": "allow" }
]
}Every Tool Call is Now Governed
Call igris.getMcpConfig() to get a signed URL for your MCP client. From this point forward, every tool call flows through Igris Sentinel — policies enforced, credentials injected, and a full audit trail recorded in Igris Lens.
// Get governed config
const config = igris.getMcpConfig(
"vk_github_prod",
{
user: "alice@company.com",
metadata: { role: "developer" },
}
);
// Use with any MCP client
const client = new McpClient({
transport: new StreamableHttp(
config.url,
{ headers: config.headers }
),
});◆ Use Cases
Built for Every Team
AI Security Teams
CISOs, Security Architects
Get complete visibility and control over every MCP tool your AI agents can reach. Set organizational RBAC policies, enforce metadata-based access control, and receive real-time alerts for policy violations.
- Prevent data exfiltration via over-permissioned MCP tools
- Role-based access control with metadata conditions
- Real-time anomaly detection and session kill switch
Platform Builders & SaaS Companies
Engineering & Product Teams
Add enterprise-grade MCP governance to the AI product you are shipping. Give your customers multi-tenant policy isolation, per-user credential management, and a complete audit trail — without building it yourself.
- Per-tenant credential vault and policy isolation
- Scoped API keys with usage metering
- White-label-ready audit trail and reporting
Compliance & Legal Teams
Compliance Officers, Legal Counsel
Maintain an auditable, immutable record of every action your AI agents take via MCP. Demonstrate that your AI systems operate within defined policies and regulatory requirements at all times.
- Full audit trail export with user attribution
- Policy version history for change management
- Compliance-ready reports via Igris Lens
◆ Get Started
Start Governing Your AI Agents Today
Igris Sentinel is available as a managed cloud service or self-hosted. Get up and running in minutes with the SDK quickstart.