Security Review Agents: 12x Fewer Tokens, 2.6x More Findings | Endor Labs

Summary

Endor Labs compared the same model, claude-sonnet-4, on 34 security prompts across 13 scopes, once with Endor Labs security evidence and once working alone from the repo and web. With the evidence, the agent used 12x fewer tokens and finished in about a third of the time. A separate benchmark found that Endor Labs AI SAST identified 2.6x more real vulnerabilities than frontier models prompted to find them on their own.

Related items

• InfoJim Cramer says Anthropic's IPO could make or break the AI trade. Here's whySame vendor · CNBC Technology
• InfoCisco AI Defense to Add Anthropic’s Claude Mythos for AI Red TeamingSame vendor · Cisco AI Blog
• InfoAnthropic releases first frontier models since Amodei's call for slowdown as IPO buzz buildsSame vendor · CNBC Technology
• InfoAnthropic launches new Claude Fable and Mythos 5.2 modelsSame vendor · The Verge (AI)
• MediumAttackers Crash Claude's 'Office Party' Flaw to Run Malicious Code on LinuxSame vendor · Dark Reading