Now live on Solana Mainnet

The kill switch
AI agents need.

Enterprise-grade safety infrastructure for autonomous AI. Real-time monitoring, instant termination, and cryptographic identity—all in under 100ms.

10

Security Modules

<100ms

Kill Latency

6,149

Lines of Code

30s

SPIFFE Revocation

One decorator.
Complete protection.

Add enterprise-grade safety to any AI agent with a single line of code. Works with OpenAI, Anthropic, LangChain, and any custom implementation.

Real-time behavioral monitoring
Automatic SPIFFE identity binding
Sub-100ms emergency termination
agent.py
from killswitch import fence

# That's it. Your agent is now protected.
@fence.protect(agent_id="prod-agent-001")
def my_ai_agent(task):
    result = llm.complete(task)
    return result

Defense in depth

10 security modules working together to ensure your AI agents stay within bounds

Runtime Fence

Monitor every action in real-time with configurable boundaries and instant alerts

Instant Kill Switch

One-click termination with SIGTERM → SIGKILL escalation for rogue agents

Audit Logging

Complete audit trail with cryptographic verification for compliance

Sub-Second Response

Kill signals processed in under 100ms when safety matters most

SPIFFE Identity

Cryptographic workload identity with automatic certificate rotation

Token Governance

Decentralized oversight with $KILLSWITCH token voting rights

$KILLSWITCH TOKEN

Stake in AI safety

Token holders govern protocol upgrades, earn subscription discounts, and participate in the decentralized oversight network.

1K+ tokens → Vote on proposals
10K+ tokens → 10% subscription discount
100K+ tokens → 20% discount + Oracle eligibility
1M+ tokens → 40% discount + 2× voting power
Buy on Jupiter

Contract Address

56o8um92XU8QMr1FsSj4nkExEkgKe56PBTAMqCAzmoon

Solana • SPL Token

Ready to secure your AI agents?

Get started in minutes. Free tier for personal projects, enterprise plans for production workloads.