Skip to main content

Integrations Overview

LangGuard integrates with AI platforms, frameworks, coding agents, and identity providers to provide unified governance and monitoring.

Integrations OverviewIntegrations Overview

Supported Integrations

AI Gateways

PlatformDescriptionStatus
OpenRouterMulti-provider AI gateway✓ Available
LiteLLMUnified LLM proxy✓ Available
CloudflareAI Gateway observability✓ Available

AI Platforms

PlatformDescriptionStatus
Azure AI FoundryDiscover AI Foundry resources, deployments, and traces✓ Available
DatabricksIngest from Unity Catalog, MLflow, Genie, and more✓ Available
AWS BedrockDiscover Bedrock models, provisioned throughput, and invocations✓ Available
Google Vertex AIGoogle Cloud AI platform🚧 Coming Soon

AI Frameworks

PlatformDescriptionStatus
MLflowML lifecycle management platform✓ Available
LangChainLLM application framework✓ Available
CrewAIMulti-agent orchestration framework✓ Available
AWS AgentCoreAmazon Bedrock AgentCore observability✓ Available

Coding Agents

PlatformDescriptionStatus
CursorAI-powered code editor integration✓ Available
Claude CodeAnthropic Claude Code integration✓ Available
OpenCodeOpen-source coding agent✓ Available
Google AntigravityGoogle AI coding agent🚧 Coming Soon

Identity Platforms

PlatformDescriptionStatus
Microsoft Entra IDIdentity governance and user enrichment✓ Available
Google WorkspaceIdentity governance and user enrichment✓ Available

Network Discovery

PlatformDescriptionStatus
SplunkSecurity and observability platform🚧 Coming Soon
Microsoft SentinelCloud-native SIEM platform🚧 Coming Soon

Endpoint Discovery

PlatformDescriptionStatus
Microsoft DefenderEndpoint detection and response🚧 Coming Soon

Architecture

LangGuard uses a unified integration architecture:

┌──────────────────────────────────────────────────────────────┐
│ LangGuard Dashboard │
├──────────────────────────────────────────────────────────────┤
│ Sync Orchestrator │
├──────────────────────────────────────────────────────────────┤
│ Connector Registry │
├─────────┬──────────┬─────────┬─────────┬─────────┬──────────┤
│ AI │ AI │ AI │ Coding │Identity │ Network │
│Gateways │Platforms │Frameworks│ Agents │Platforms│Discovery │
└─────────┴──────────┴─────────┴─────────┴─────────┴──────────┘

Key Features

  • Unified Credential Storage - AES-256-GCM encrypted credentials
  • Automatic Sync - Configurable polling intervals
  • Push-Based Ingestion - OpenTelemetry webhook support
  • Source Tracking - Visual indicators for data origin
  • Extensible - Easy to add new integrations

Adding an Integration

Via UI

  1. Navigate to Integrations in the sidebar
  2. Click Add Integration
  3. Browse categories to find your platform
  4. Click the platform card
  5. Enter credentials
  6. Test connection
  7. Save

Credential Security

Encryption

All credentials are encrypted before storage:

  • Algorithm: AES-256-GCM
  • Key: 256-bit key from ENCRYPTION_KEY env var
  • Storage: Only encrypted data in database
  • Transit: HTTPS for all API calls

Best Practices

  1. Use service accounts - Don't use personal API keys
  2. Rotate regularly - Update keys quarterly
  3. Least privilege - Only grant needed permissions
  4. Monitor access - Review API key usage logs

Sync Configuration

Automatic Sync

Configure automatic synchronization:

SettingDescriptionDefault
EnabledAuto-sync on/offYes
IntervalTime between syncs15 minutes
LookbackDays of history7 days
Batch SizeItems per sync100

Manual Sync

Trigger sync on-demand by clicking the Sync Now button on the integration card.

Push-Based Ingestion

For real-time ingestion, use OpenTelemetry:

POST /v1/traces
Headers:
x-tenant-slug: your-tenant
x-otel-signature: sha256=<hmac-signature>

See individual integration guides for setup.

Monitoring Integration Health

Connection Status

StatusMeaning
🟢 ConnectedCredentials valid, sync working
🟡 WarningMinor issues (rate limits, etc.)
🔴 FailedConnection or auth failed
⚪ DisabledIntegration manually disabled

Sync History

View recent sync operations:

  • Success/failure status
  • Items fetched and ingested
  • Duration
  • Error messages (if any)

Troubleshooting

Common issues and solutions:

  1. Auth Failed - Check credentials are correct
  2. Rate Limited - Increase sync interval
  3. Timeout - Reduce batch size
  4. No Data - Check time range and filters

See Integration Issues for detailed help.

Integration Guides


Need a New Integration?

Don't see your platform? Contact info@langguard.ai to request an integration.