Garak vs NeMo Guardrails
NeMo Guardrails leads on trust: 71.9/100 (Grade B) against 55.8/100 (Grade C), a 16.1-point gap. Garak leads on maintenance; NeMo Guardrails leads on supply-chain integrity and transparency.
the LLM vulnerability scanner
Choose Garak if maintenance matters most.
- +0.9Maintenance
NeMo Guardrails is an open-source toolkit for easily adding programmable guardrails to LLM-based conversational systems.
Choose NeMo Guardrails if supply-chain integrity and transparency matter most.
- +4.6Safety / Integrity: 100% of recent commits signed, against 68%
- +2.0Transparency: OSSF Scorecard 7.1 against 4.7
- +1.1Adoption: 89.7k weekly downloads against 8.7k
Where they differ
1 dimension identical: Identity 10.8 · Full evidence table
An independent, evidence-based trust comparison of Garak and NeMo Guardrails, two Security & Guardrails projects in the HVTracker registry. Scores come from public, checkable signals — supply-chain provenance, OSSF Scorecard, maintenance, and adoption — not popularity.
Full evidence
| Signal | Garakleondz/garak | NeMo GuardrailsNVIDIA/NeMo-Guardrails |
|---|---|---|
| HVTrust score | 55.8 | 71.9 |
| Evidence grade | C | B |
| Coverage grade | A | A |
| Overall rank | #759 | #290 |
| Rank in Security & Guardrails | #13 | #5 |
| GitHub stars | 9.4k | 7.2k |
| Last updated | 13d ago | 1d ago |
| Build provenance | No | No |
| OSSF Scorecard | 4.7 / 10 | 7.1 / 10 |
| License | Apache-2.0 | NOASSERTION |
| Downloads | 9k/wk | 90k/wk |
| Trust dimensions (points earned) | ||
| Safety / integrity / 25 | 9.3 | 13.9 |
| Identity & provenance / 18 | 10.8 | 10.8 |
| Transparency / 17 | 12.5 | 14.5 |
| Maintenance / 20 | 17.7 | 16.8 |
| Adoption / 20 | 14.8 | 15.9 |
| Runtime capability surface (full matrix) | ||
| MCP server | — | — |
| External providers | 8 — Amazon Bedrock, Anthropic, Cohere, … | 1 — OpenAI |
| Requires API keys | Yes | No |
| Plugin surface | plugins | — |
| Provenance drift | Warning | Unknown |
How to read this: HVTrust (0–100) weighs supply-chain signals (provenance, OSSF Scorecard, signed commits, open license) alongside real-world adoption, scaled by an evidence-confidence factor. Grade bands: A ≥ 80, B ≥ 65, C ≥ 50, D < 50. Signals refresh daily. Full methodology v4.3 →