---
name: "nemo-guardrails-dialogue-rails"
description: "Steer dialogue flows and enforce factual boundaries using NVIDIA NeMo Guardrails."
metadata:
  author: "NVIDIA"
  version: "1.0.0"
  runtime_compatibility:
    - "codex"
    - "claude-code"
    - "cursor"
    - "windsurf"
triggers:
  - "setup nemo guardrails"
  - "write colang rails"
  - "enforce on-topic guardrails nvidia"
permissions:
  fs_read: true
  fs_write: "workspace"
  shell_exec: false
  net_access: false
---

# NVIDIA NeMo Guardrails Topic & Factuality Rails

Configures NeMo Guardrails using Colang (.co) syntax: topic control (keeping conversations on topic), fact-checking rails, hallucination mitigation, and jailbreak prevention.

## When to Invoke
Invoke this skill whenever you need to:
- setup nemo guardrails
- write colang rails
- enforce on-topic guardrails nvidia

## Execution SOP & Rules
1. Configure config.yml declaring models, rails, and instructions.
2. Author Colang flows in .co files defining user intents, bot responses, and flow triggers.
3. Enable self_check_facts rail to prevent hallucinated factual assertions.
4. Test dialogue flows with automated conversational test suites to verify guardrails efficacy.

## Verification Checklist
- Run local linting, syntax validation, and type-checks before reporting completion.
- Ensure all created or modified files adhere to workspace boundaries and avoid uncommitted secrets.
- Verify tests or build commands exit cleanly with return code 0.