Kvlar SHIELD

Enterprise agent
security.

Managed runtime enforcement for AI agents. Define policies as code, enforce them at the transport layer, and audit every action. Built on the open-source Kvlar SDK.

Request a demo→Documentation

Pricing

Pick your plan.

Community

Free

For individuals exploring agent security.

  • Agents3
  • Actions / month10K
  • Team members1
  • Policies5
  • Audit retention7 days
  • SupportCommunity
Start Free

Pro

$49/mo

For developers shipping agents in production.

  • Agents10
  • Actions / month100K
  • Team members3
  • Policies25
  • Audit retention30 days
  • SupportEmail
Get Started

Team

$199/mo

For teams that need control, compliance, and visibility.

  • Agents50
  • Actions / month1M
  • Team members10
  • PoliciesUnlimited
  • Audit retention90 days
  • RBACIncluded
  • SIEM exportIncluded
  • SLA99.9%
  • SupportPriority
Get Started

Enterprise

Custom

For organizations with advanced security requirements.

  • AgentsUnlimited
  • Actions / monthUnlimited
  • Team membersUnlimited
  • PoliciesUnlimited
  • Audit retentionCustom
  • RBACIncluded
  • SIEM exportIncluded
  • SSO / SAMLIncluded
  • SLA99.99%
  • Dedicated supportIncluded
  • Compliance reportsIncluded
Contact Sales

Beyond the SDK

What SHIELD
adds.

The open-source Kvlar SDK is the policy engine. SHIELD wraps it in a managed control plane with team features, audit infrastructure, and enterprise integrations.

Centralized policies

Push policy updates to all agents from a single control plane. No redeployments, no config drift.

Role-based access

RBAC for policy authoring, approval workflows, and audit access. Enforce separation of duties.

Audit & compliance

Structured audit trail with retention guarantees. Export to Splunk, Datadog, or any SIEM via CEF/JSONL.

Approval routing

Route require_approval actions to Slack, PagerDuty, or custom webhook endpoints with configurable timeouts.

Real-time dashboard

Live view of agent actions, policy decisions, and security posture across your entire fleet.

Enterprise SSO

SAML 2.0 and OIDC integration for single sign-on. Enforce identity policies across your organization.

Comparison

Full feature
breakdown.

FeatureCommunityProTeamEnterprise
Usage
Agents
3
10
50
Unlimited
Actions / month
10K
100K
1M
Unlimited
Team members
1
3
10
Unlimited
Policies
5
25
Unlimited
Unlimited
Security
Policy engine
Fail-closed enforcement
Approval webhooks
RBAC
SSO / SAML
Policy hot-reload
Observability
Audit logging
7 days
30 days
90 days
Custom
SIEM export (CEF, JSONL)
Real-time dashboard
Anomaly detection
Support & Compliance
Community support
Email support
Priority support
Dedicated support engineer
SLA
—
—
99.9%
99.99%
Compliance reports

FAQ

Common
questions.

What happens if I exceed my action limit?

We won't cut off your agents mid-session. You'll get a notification and a 7-day grace period to upgrade or reduce usage. After that, actions above your limit will be queued for review.

Can I use the open-source SDK without SHIELD?

Absolutely. The Kvlar SDK is Apache 2.0 licensed and fully functional on its own. SHIELD adds managed infrastructure, team features, and enterprise integrations on top.

How does SHIELD differ from the SDK?

The SDK is the policy engine that runs locally. SHIELD adds a centralized control plane: push policies to all agents, manage team access with RBAC, retain audit logs with SLA guarantees, and export to your SIEM.

Do you offer annual billing?

Yes. Annual plans get two months free. Contact sales for details.

What compliance standards does SHIELD support?

Enterprise plans include compliance reports for SOC 2 Type II, GDPR, and HIPAA. We can work with your compliance team on specific requirements.

Can I self-host SHIELD?

Enterprise plans include a self-hosted deployment option. Contact sales for architecture details and requirements.

Secure your agents
before they ship.

Start with the free Community tier. Upgrade when your team grows. Enterprise plans include dedicated support and compliance.

Start free→Contact sales