SAILLENT Documentation

The complete guide to deploying, configuring, and operating SAILLENT - the runtime security platform for AI agents, MCP servers, and LLM applications.

Overview

SAILLENT is a runtime security and governance platform that intercepts, inspects, and enforces security policies on every AI action before execution. We provide the critical infrastructure layer that sits between your AI applications and the models, tools, and data they access.

The Problem We Solve

Traditional security tools were not built for AI:

Firewalls do not understand prompts.
WAFs cannot detect prompt injection.
SIEMs cannot track autonomous agent behavior.

As enterprises deploy AI agents at scale, they face critical challenges: no visibility into agent behavior, no control over tool access, no audit trail for compliance, and no governance framework for AI security policies.

The SAILLENT Solution

SAILLENT intercepts every AI action in real-time - prompts, tool calls, and autonomous decisions. We inspect, evaluate against policies, and enforce security controls before any action executes.

Key Capabilities:

Real-time inspection of all AI interactions
Policy-based enforcement with fine-grained control
Complete audit trails for compliance and forensics
Prompt injection detection to prevent manipulation
PII redaction to protect sensitive data
MCP authorization for tool access control
SIEM integration for security operations

Architecture

SAILLENT operates as a distributed runtime inspection layer with centralized policy management and edge-deployed enforcement engines.

Control Plane

Centralized management for policy definition, configuration, and monitoring.

Runtime Engine

High-performance enforcement engines deployed at the edge for sub-millisecond inspection.

Policy Engine

Evaluates AI actions against security policies with context-aware decisions.

Audit Engine

Records every decision with complete context for compliance and forensics.

Quick Start

Get SAILLENT up and running in under 5 minutes.

Step 1: Install SAILLENT

# Pull the latest image
docker pull saillent/runtime-engine:latest

# Start the runtime engine
docker run -d \
  --name saillent-engine \
  -p 8080:8080 \
  -p 8443:8443 \
  -e SAILLENT_API_KEY=<your-api-key> \
  -e SAILLENT_CONTROL_PLANE=<your-control-plane-url> \
  saillent/runtime-engine:latest

Step 2: Create Your First Policy

Create block-pii.yaml:

policy: "block-pii-leakage"
description: "Block requests containing PII"
action: "block"
priority: 100
conditions:
  - type: "pii_detection"
    entities:
      - "SSN"
      - "CreditCard"
      - "EmailAddress"
enforcement: "block"
audit: true

Policy Engine

The Policy Engine evaluates every AI action against your security policies. Policies are written in YAML and consist of:

Name: Unique identifier
Action: Enforcement action (allow, block, redact, require-approval, log-only)
Priority: Evaluation order (higher = first)
Conditions: Matching criteria

Best Practices:

Start with log-only policies to understand your environment
Gradually move to enforcement as you gain confidence
Use descriptive policy names and comments

MCP & LLM Security

SAILLENT provides comprehensive security for Model Context Protocol (MCP) communications and LLM interactions.

Protection Capabilities

Tool Authorization: Control which tools each agent can access
Parameter Inspection: Validate all parameters in tool calls
Prompt Injection Detection: Block malicious prompts that manipulate model behavior
PII Redaction: Automatically redact sensitive information from prompts and responses

API Reference

SAILLENT provides a REST API for managing policies, retrieving audit logs, and monitoring system health.

# Authentication
Authorization: Bearer <your-api-key>

# Health Check
GET /health

# List Policies
GET /api/v1/policies

# Create Policy
POST /api/v1/policies
Content-Type: application/yaml

# Get Audit Logs
GET /api/v1/audit/logs?limit=100&offset=0

Troubleshooting

Engine Won't Start

Solution: Check logs (docker logs saillent-engine), verify API key, ensure Control Plane URL is reachable, and check port availability.

Policies Not Applying

Solution: Verify policy is enabled via API, check policy priority, ensure conditions match request context, and review audit logs for evaluation details.

High Latency

Solution: Check network latency to Control Plane, reduce active policies, optimize regex conditions, and scale horizontally with more engine instances.

Need detailed technical documentation?

Contact Support