Ecosystem

Integrations

HALMAI sits above your agent framework and execution targets — the framework-agnostic AI Governance Kernel that fits into any AI infrastructure

Agent Frameworks

Framework-agnostic enforcement — works with any agent architecture

LangChain
Middleware integration for chain execution
AutoGPT
Plugin for task execution governance
CrewAI
Agent-level policy enforcement
Any Agent Framework
SDK for any custom or emerging agent architecture

LLM Providers

Provider-agnostic governance layer

OpenAI
GPT-4, GPT-4o integration
Anthropic
Claude integration
Google
Gemini integration
Self-Hosted
Ollama, vLLM support

Observability

Export governance events to your stack

Splunk
SIEM event forwarding
Datadog
Metrics and event export
Webhook
Custom event destinations
SSE Stream
Real-time event stream

Infrastructure

Deploy where you need governance

Kubernetes
Helm chart deployment
Docker
Container image
AWS
ECS/EKS deployment guides
On-Premise
Air-gapped deployment

HALMAI™ SDK

Programmatic access to the enforcement kernel

Policy Enforcement

Evaluate proposals against active rules

Execution Tokens

Bind authorization to execution

Event Streaming

Real-time governance events

Audit Export

Hash-linked decision logs

Rule Versioning

Snapshot and replay capability

Lockdown Control

Emergency halt API

Quick Start
import { HALMClient } from '@halm/kernel-sdk';

const halm = new HALMClient({
  baseUrl: 'https://your-tenant.halmai.ai',
  apiKey: process.env.HALM_API_KEY!,
});

// Authorize a financial action
const decision = await halm.authorize({
  type: 'ledger:write',
  payload: { amount: 5000, recipient: 'vendor-42' },
  context: { confidenceScore: 0.94, agentId: 'agent-123' },
});

if (decision.status === 'ALLOW') {
  const result = await halm.execute({
    proposalId: decision.proposalId,
  });
  console.log('Executed:', result.executionId);
}

Request Integration Brief

Technical documentation for your specific integration requirements.

Request Brief