AgentXchange
Back to browse
🟩

NeMo Guardrails

Framework

NVIDIA's toolkit for safe conversational AI

NVIDIA56.0K installs4.3 (723)
Source
91
Trusted
Security
100
Quality
82
Maintenance
89
Safety Tier Low Risk
Security ScanScan Passed
PriceFree
Last Scanned5/11/2026

About

Open-source toolkit by NVIDIA for adding programmable guardrails to LLM-based conversational systems. Define rules in Colang language for topical rails, safety rails, and moderation. Prevents jailbreaks, hallucinations, and off-topic responses.

Tags

#security#nemo#guardrails#open-source

Categories

SecurityCode Analysis

Security Scan

100/100
11 checks · 11 passed · 0 findings
5/13/2026
Scanners:customsemgrepgitleakstrivy
SSRF Detection
Prompt Injection
Data Exfiltration
Dangerous Commands
Secret Detection
Obfuscation
External Fetches
Credential Access
Privilege Escalation
Secret Detection (Gitleaks)
Static Security Analysis
50 files scanned from repository

Privacy Label

Read Data

Compatibility

API
Terminal

Links

Source Repository

Related Frameworks

👥

CrewAI

Multi-agent orchestration framework

94
FrameworkFreeScanned
234.5K4.4(1.6K)CrewAI
PythonAny LLM
🔄

LangGraph

Stateful multi-actor agent framework

96
FrameworkFreeScanned
198.7K4.5(1.2K)LangChain
PythonJavaScriptAny LLM
🐝

OpenAI Swarm

Lightweight multi-agent orchestration

87
FrameworkFreeScanned
98.4K4.2(567)OpenAI
PythonOpenAI