Security-One: open-weight 27B model outputs probabilities for agent security decisions
huggingface · x · 2026-10-05
Red Hat AI released Security-One, an open-weight 27B model built for security decisions.
- It ingests a document, tool call, or code change
- It returns probabilities for answers you define, rather than hard verdicts
- Your own code decides what to allow, block, or route to human review based on those probabilities
The design separates probability estimation from policy enforcement, making it a natural guardrail inside agent workflows. Weights are available on Hugging Face.
More from coding & agent
- pg-jev open-source Postgres extension runs semantic filters and scoring inside SQL — Arindam_1729 · 2026-10-06
- Open-source Discord AI assistant Zauq v4 ships bounded agents, MCP and sandboxed code verification — rar_file-exe · 2026-10-06
- Orbio launches new agent tools: sandboxes, 24/7 servers, databases and email, paid in CREDIT — econoar · 2026-10-06
- New OSWorld leader: 79.38% success at $14.34/task, 40% cheaper than Claude — xwang_lk · 2026-10-06
- Job hunting with AI: build skills weekly, then out-hustle recruiters on DMs — brandon_galang · 2026-10-06
- 'Open is the way': dev lists 10 notable Open-named tools — rajistics · 2026-10-06