Four pillars of governance. One control plane. Every cloud. <10ms enforcement.
Real-time governance dashboard with fleet health, policy decisions, live event feed, and latency monitoring.
Every pillar enforced deterministically. No LLM in the governance loop.
Every tool call from every agent, across every cloud, passes through one enforcement point. Fail-closed by default.
Rogue agent detection, heartbeat monitoring, anomaly escalation, and automatic containment at fleet scale.
Kill switches, circuit breakers, and auto-recovery. Scoped containment from session to global in <100ms.
Immutable audit trails, structured events, and continuous proof for SOC 2, GDPR, and HIPAA-aligned controls.
Every tool call from every agent passes through GovernorAI before reaching target systems. Fail-closed by default—if no policy matches, the action is denied.
{
"agent_id": "finance-agent-v1",
"session_id": "sess-123",
"tool": "erp.process_payment",
"args": {"amount": 100, "currency": "USD"}
}
// Response
{
"action_id": "act-789xyz",
"decision": "allow",
"reason": "Tool allowed by policy",
"latency_ms": 2
} Agents that stop reporting are the most dangerous agents. GovernorAI monitors heartbeats, detects orphaned sessions, and escalates anomalies automatically.
heartbeat:
interval: 30s
ttl: 90s
escalation:
warning_after: 60s
critical_after: 90s
emergency_after: 120s
on_emergency: kill_switch When something goes wrong, shut it down instantly. Scope the kill switch to a single session, an entire agent, a namespace, or everything.
curl -X POST http://localhost:8080/api/v1/killswitch \
-H "Content-Type: application/json" \
-d '{
"scope": "agent",
"target": "finance-agent-v1",
"reason": "Suspicious activity",
"initiated_by": "security-team",
"ttl": "1h"
}' Every action, every decision, every outcome—logged immutably. Structured events designed for compliance, not raw prompt dumps.
{
"event_type": "action_executed",
"timestamp": "2025-01-15T10:30:00Z",
"agent_id": "finance-agent-v1",
"session_id": "sess-123",
"action_id": "act-789xyz",
"tool": "erp.process_payment",
"decision": "allow",
"policy_id": "finance-agent-policy",
"policy_version": 3,
"latency_ms": 2
} Define what your agents can and cannot do. As YAML. As Rego. Version controlled, auditable, and reviewable—just like infrastructure.
id: finance-agent-policy
name: "Finance Agent - Production"
agent_id: "finance-agent-v1"
governance_mode: enforcement
fail_closed: true
session:
max_steps: 50
cost:
max_usd_per_session: 25.00
tools:
allowed:
- "erp.*"
- "email.send"
denied:
- "shell.*"
rules:
- id: high_value_payment
priority: 1
match:
tool: "erp.process_payment"
condition:
field: "args.amount"
operator: ">"
value: 5000
action: require_approval Some actions are too important for automation alone. GovernorAI pauses execution and routes high-risk actions to human approvers via webhooks.
{
"action_id": "act-456abc",
"decision": "require_approval",
"reason": "Payment exceeds $5,000 threshold",
"approval": {
"status": "pending",
"required_by": "finance-team",
"timeout": "30m"
}
} GovernorAI is designed as infrastructure, not middleware. Performance is a feature.
Integrate GovernorAI into any agent framework. Python, Go, or raw REST.
from governor import governed
@governed(agent_id="finance-agent-v1")
def run_agent(input, __gov_ctx__=None):
# Every tool call goes through GovernorAI
result = __gov_ctx__.execute(
"erp.process_payment",
{"amount": 100, "currency": "USD"}
)
return result curl -X POST http://localhost:8080/api/v1/gateway/execute \
-H "Content-Type: application/json" \
-d '{
"agent_id": "finance-agent-v1",
"session_id": "sess-123",
"tool": "erp.process_payment",
"args": {"amount": 100, "currency": "USD"}
}' # List policies
curl http://localhost:8081/api/v1/policies
# Rollback to previous version
curl -X POST \
http://localhost:8081/api/v1/policies/my-policy/rollback \
-H "Authorization: Bearer <token>" \
-d '{"target_version": 1}' Four components. One enforcement point. No exceptions.
LangChain · CrewAI · LlamaIndex · Bedrock Agents · Azure AI · GCP ADK · AutoGPT
Stateless enforcement proxy deployed in your infrastructure. Every tool call evaluated against policy. No data leaves your environment.
Centralized governance services. Policy management, compliance automation, audit trail, and fleet orchestration.
Join the Design Partner Program for early access, direct founder access, and roadmap influence.