Last updated · 2026-06-23

Acceptable Use Policy

This Acceptable Use Policy ("AUP") forms part of our Terms of Service and applies to everything you run on Superjolt — including anything an AI agent does on your behalf. The goal is simple: keep the platform safe for everyone and keep us in good standing with our upstream providers. Breaching this policy is a breach of the Terms.

Prohibited uses

You must not use Superjolt to:

  • Send spam or unsolicited bulk email. No spam, phishing, or unsolicited bulk mail, and no infrastructure that supports it. Direct outbound SMTP from VMs is blocked at the network layer (ports 25, 465, 587, and 2525), and the shared sandbox sending domain is rate-capped; legitimate mail should go through a verified sending domain.
  • Scan, probe, or attack third-party systems. No port-scanning, vulnerability scanning, brute-forcing, or other unsolicited probing of systems you don't own. Per-VM connection-rate and concurrent-connection limits are enforced to contain this.
  • Launch denial-of-service or amplification attacks. No DoS/DDoS, reflection, or amplification traffic. Known amplification UDP ports are blocked and per-VM egress bandwidth is shaped.
  • Attack Superjolt or other customers. No attempts to break tenant isolation, reach another customer's VMs, spoof addresses, or access platform internals or cloud metadata endpoints. These boundaries are enforced in the network layer.
  • Run malware or credential-abuse infrastructure. No malware, botnet command-and-control, or credential-stuffing operations.
  • Mine cryptocurrency. Crypto mining is not permitted.
  • Host illegal content. No CSAM and nothing that is illegal in the jurisdictions where you or we operate.
  • Resell raw capacity. Don't resell raw Superjolt compute to others as if it were your own platform.
  • Violate AI usage policies. When the platform is driven by an AI agent, don't use it for anything the relevant AI provider's usage policies (e.g. Anthropic's) prohibit.

How we enforce it

We don't read the contents of your VMs. We enforce this policy by signal — abuse reports, network traffic patterns, and billing-fraud signals — combined with the network-layer controls above. The technical limits (blocked ports, connection-rate caps, egress shaping) apply automatically to every VM; they're a backstop, not a substitute for following this policy.

Consequences

If you breach this policy we may throttle, suspend, or terminate the offending workload or your account. Where the situation allows, we'll reach out first and give you a chance to fix it. Severe or ongoing abuse — active attacks, spam runs, or illegal content — gets an immediate cut to protect the platform and other customers.

Reporting abuse

Seeing abuse from a Superjolt IP or domain? Email [email protected] with the offending IP / domain and any logs, and we'll act fast.