Secures AI agents and apps at runtime: detects prompt injection, enforces tool-call policies, redacts sensitive data, and blocks bots, through SDKs, an MCP server, and a CLI.
- Access
- APIMCPCLI
- Auth
- API Key
- Skills
- 5
Skills
Prompt Injection Detection
Detects prompt injection attempts in prompts and untrusted input before an agent or model acts on them.
Guard Policies
Authorizes agent tool calls and actions against policies written in Rego, with drafting, validation, and test evaluation.
Sensitive Info Detection
Detects sensitive information such as PII in requests and agent traffic so it can be blocked or redacted.
Bot Protection
Identifies and blocks automated clients and abusive traffic on HTTP routes, API endpoints, and agent entry points.
Request Investigation
Explains why a request was allowed or denied and analyzes traffic, anomalies, and IP threat intelligence for a site.
Related Agents
NVIDIA NemoClaw
CLIOpen-source security runtime for multi-agent AI. Enforces policy-based access control, sandboxes kernel interactions…
NeMo Guardrails
APIAdds programmable guardrails to LLM apps — input/output moderation, jailbreak and prompt-injection detection, topic…
Kong AI Gateway
API · CLI · MCPGoverns and meters LLM and agent traffic through a high-performance API gateway, with an MCP gateway that turns…
HOL Guard
CLIBlocks risky AI coding agent actions at runtime on your machine — shell commands, secret access, MCP servers, plugins…