Enterprise Service Level Agreement (SLA)
This Service Level Agreement (“SLA”) governs the performance, support availability, incident response timelines, and uptime commitments provided by RitualDev Lab for enterprise deployments of AgentShield, DBShield, and related governance tooling.
01.Scope & Coverage
This SLA is incorporated into and forms an integral part of the Enterprise Master Services Agreement (MSA) between RitualDev Lab (“Vendor”) and the enterprise customer (“Customer”). This agreement applies to:
- AgentShield Autonomous Sandbox (Kubernetes Helm deployments, Docker daemons, and workstation CLI runtimes).
- DBShield Wire Guardrail (Database mutation proxies and MCP server instances).
- Enterprise Governance Suite (SIEM exporters, telemetry webhooks, and private VPC infrastructure modules).
02.Incident Severity Classification & Response Guarantees
When an enterprise incident ticket is logged via our designated dispatch channels, RitualDev commits to the initial triage and ongoing status cadence defined below:
Critical Outage / Zero-Day
Total sandbox bypass in active production agent workflows, zero-day kernel/AST escape vulnerability, or verified customer credential exfiltration.
Partial Degradation
Core security feature or AST normalizer false-positive halting continuous deployment CI/CD pipeline, SIEM exporter failure without security bypass.
General Guidance
Non-critical inquiry, configuration assistance, bespoke regex rule tuning, architecture integration advice, or general roadmap queries.
03.Uptime & Air-Gapped High Availability
Because AgentShield is engineered as a strictly local-first, air-gapped security boundary, runtime execution availability is not dependent upon third-party cloud infrastructure.
Guaranteed engine operational availability for containerized Helm daemonsets running inside Customer VPCs.
In the rare event of daemon termination, subshell interception locks by default to prevent unprotected command execution.
04.Multi-Tier Engineering Escalation Ladder
Initial diagnostic triage, reproduction sandbox validation, and log intake.
AST parser patching, linear-time regex optimization, and live emergency hotfix deployment.
Executive coordination, emergency private patch signing, and joint incident post-mortem drafting.
05.Service Credits & Remedies
In the event that RitualDev fails to meet the guaranteed P1 or P2 initial response SLA for an Enterprise customer within a given calendar month, Customer shall be entitled to service credits calculated as follows:
| SLA Breach Metric | Service Credit (% of Monthly Fee) |
|---|---|
| P1 Response Time exceeds 1 hour (up to 3 hours) | 10% Credit |
| P1 Response Time exceeds 3 hours | 25% Credit |
| Monthly Containerized Daemon Availability < 99.9% | 25% Credit |
| Monthly Containerized Daemon Availability < 99.0% | 50% Credit |
Service credits are applied against the customer's next billing cycle and represent the sole and exclusive financial remedy for SLA response breaches.
06.SLA Exclusions
SLA commitments and remedies do not apply to service interruptions or response delays resulting from:
- Customer-managed infrastructure failures, cloud VPC network disconnects, or operating system kernel panics outside of Vendor container limits.
- Unauthorized alterations, reverse engineering, or modifications to AgentShield binary distributions.
- External upstream third-party LLM cloud outages (e.g. Anthropic Claude, OpenAI, or Microsoft Azure API downtime).
- Force majeure events beyond reasonable commercial control.
Authorized Enterprise Dispatch Channels
To trigger an SLA response, incidents must be dispatched to the verified addresses below with the corresponding priority indicator in the email subject line: