6 open-source agents tracked · Ranked by trust score · Updated 2026-05-30 20:03 UTC
HVTracker independently evaluates 6 open-source security & guardrails using daily signals from GitHub, package registries, and security databases. Each agent is scored on activity, adoption, transparency, safety, and identity. The top-ranked security & guardrails is NeMo Guardrails with a trust score of 71.8/100 (Grade B). Other leading projects include Garak and Guardrails AI.
| # | Agent | Trust | Stars | Language |
|---|---|---|---|---|
| 1 | NeMo Guardrails B NeMo Guardrails is an open-source toolkit for easily adding programmable guardrails to LLM-based conversational systems. | 71.8 | 6.3k | Python |
| 2 | Garak B the LLM vulnerability scanner | 67.0 | 8.0k | Python |
| 3 | Guardrails AI B Adding guardrails to large language models. | 61.4 | 6.9k | Python |
| 4 | Llama Guard C Set of tools to assess and improve LLM security. | 54.3 | 4.2k | Python |
| 5 | LLM Guard B The Security Toolkit for LLM Interactions | 48.9 | 3.0k | Python |
| 6 | Rebuff B LLM Prompt Injection Detector | 28.6 | 1.5k | TypeScript |