# GenAI Protector Plus

> Runtime AI security for GenAI and agentic AI workloads — official documentation, deployment guides, and API reference for CloudsineAI's GenAI Protector Plus.

## Docs

- [Authentication](https://docs.protectorplus.cloudsine.tech/api-reference/authentication.md): Generate and use your Protector Plus API key.
- [Guardrails Reference](https://docs.protectorplus.cloudsine.tech/api-reference/guardrails-reference.md): All guardrails surfaced by the inspection APIs.
- [health](https://docs.protectorplus.cloudsine.tech/api-reference/health.md): Unauthenticated liveness probe.
- [input-check](https://docs.protectorplus.cloudsine.tech/api-reference/input-check.md): Inspect a user prompt before it reaches the LLM.
- [Integration Pattern](https://docs.protectorplus.cloudsine.tech/api-reference/integration-pattern.md): Wire input-check and output-check into your LLM call path.
- [API Reference](https://docs.protectorplus.cloudsine.tech/api-reference/introduction.md): Inspect prompts and responses through Protector Plus over HTTP.
- [output-check](https://docs.protectorplus.cloudsine.tech/api-reference/output-check.md): Inspect an LLM response before it returns to the user.
- [Test Prompts](https://docs.protectorplus.cloudsine.tech/api-reference/test-prompts.md): Ready-to-run prompts that exercise each guardrail.
- [Deployment Modes](https://docs.protectorplus.cloudsine.tech/architecture/deployment-modes.md): Where the firewall sits, who blocks, and how applications adopt enforcement.
- [Detection Layers](https://docs.protectorplus.cloudsine.tech/architecture/detection-layers.md): Standard Guardrails and ShieldPrompt — two complementary layers in defence-in-depth.
- [ShieldPrompt](https://docs.protectorplus.cloudsine.tech/architecture/shieldprompt.md): Defence-in-depth runtime pipeline combining classifier, vector, content-moderation, and system-prompt detection.
- [System Overview](https://docs.protectorplus.cloudsine.tech/architecture/system-overview.md): The runtime inspection layer between your GenAI application and the LLM.
- [Threat Vector Database](https://docs.protectorplus.cloudsine.tech/architecture/threat-vector-database.md): CloudsineAI's proprietary corpus of LLM attack patterns, maintained continuously by the CloudsineAI research operation.
- [Accuracy](https://docs.protectorplus.cloudsine.tech/benchmarks/accuracy.md): Latest accuracy benchmarks for Protector Plus across PII, prompt injection, system-prompt protection, and unsafe content.
- [Performance](https://docs.protectorplus.cloudsine.tech/benchmarks/performance.md): Latency and throughput on the recommended hardware specification.
- [Test Harness & Methodology](https://docs.protectorplus.cloudsine.tech/benchmarks/uat-methodology.md): How CloudsineAI measures Protector Plus accuracy — the eval harness, the corpora, and what each test exercises.
- [AWS Marketplace AMI](https://docs.protectorplus.cloudsine.tech/deployment/aws-marketplace.md): Deploy Protector Plus from the AWS FTR-certified Amazon Machine Image.
- [Kubernetes / Helm](https://docs.protectorplus.cloudsine.tech/deployment/kubernetes-helm.md): Enterprise deployment via Helm chart for multi-replica scale and CD pipelines.
- [Network Prerequisites](https://docs.protectorplus.cloudsine.tech/deployment/network-prerequisites.md): Ports, protocols, and DNS for a Protector Plus deployment.
- [On-Premises & Air-Gapped](https://docs.protectorplus.cloudsine.tech/deployment/on-premises.md): Sovereign and air-gapped deployment as a first-class mode.
- [Key Concepts](https://docs.protectorplus.cloudsine.tech/introduction/key-concepts.md): The building blocks of Protector Plus — guardrails, ShieldPrompt, security profiles, and the TVDB.
- [Overview](https://docs.protectorplus.cloudsine.tech/introduction/overview.md): Runtime AI security for GenAI and agentic AI — production-deployed in Singapore.
- [Quickstart](https://docs.protectorplus.cloudsine.tech/introduction/quickstart.md): Deploy Protector Plus from AWS Marketplace, create your first security profile, and validate a request end-to-end.
- [Account & Role Management](https://docs.protectorplus.cloudsine.tech/user-guide/account-management.md): Local accounts, Enterprise SSO, and the role-based access matrix.
- [Configuring Guardrails](https://docs.protectorplus.cloudsine.tech/user-guide/guardrails.md): Per-guardrail enablement, block/monitor mode, and worked configuration examples.
- [Initial Setup](https://docs.protectorplus.cloudsine.tech/user-guide/initial-setup.md): First-time provisioning of the Protector Plus appliance and dashboard.
- [Protection Levels](https://docs.protectorplus.cloudsine.tech/user-guide/protection-levels.md): Templated guardrail presets — Level 0 through Level 3 and Full Custom.
- [Security Profiles](https://docs.protectorplus.cloudsine.tech/user-guide/security-profiles.md): Bind GenAI applications to packaged guardrail configurations.
- [SIEM Integration](https://docs.protectorplus.cloudsine.tech/user-guide/siem-integration.md): Forward alerts and dashboard events to your log collector.
- [Threat Management](https://docs.protectorplus.cloudsine.tech/user-guide/threat-management.md): Alerts, dashboards, and forwarding to your SIEM.

## OpenAPI Specs

- [openapi](https://docs.protectorplus.cloudsine.tech/api-reference/openapi.json)