Menu
Verify response faithfulness against RAG source documents using NLI entailment and LLM-as-judge
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Trust-aware dialogue generation for AI companion relationships
Generate, edit, upscale, variate, and style-transfer images using the AgentOS multi-provider image pipeline with automatic fallback and character consistency.
Difficulty scaling, CR calculations, action economy, adaptive encounter design
Fair adjudication of game encounters — combat, puzzles, social challenges
Renders game state changes into dramatic, immersive prose narrative
Branching quest DAGs, prospective memory triggers, workflow composition
| name | grounding-guard |
| version | 1.0.0 |
| description | Verify response faithfulness against RAG source documents using NLI entailment and LLM-as-judge |
| author | Frame.dev |
| namespace | wunderland |
| category | security |
| tags | ["guardrails","hallucination","grounding","faithfulness","nli","rag","fact-checking"] |
| requires_tools | ["check_grounding"] |
| metadata | {"agentos":{"emoji":"🔍"}} |
A guardrail automatically verifies that your responses are faithful to the source documents retrieved via RAG. Claims in your output are checked against the retrieved sources using NLI entailment detection.
When HITL is enabled with guardrailOverride: true, the grounding guard participates in the post-approval guardrail pipeline. If a tool action is approved (by human or LLM judge) but produces output containing ungrounded claims, the grounding guard can flag or veto the response before it reaches the user. See the hitl-safety skill for HITL configuration details.
