Evidence · DiggingBeagle record
Last Secure Code Benchmark finds large functionality-security gap in AI coding agents — original research or disclosure
Reported evidence for A 450-task study reports much higher functional completion than joint functionality and security success across six coding agents. Research intake: DB-RADAR-CP7-CP7A-002.
- Source role
- primary disclosure
Evidence record
Reported evidence for A 450-task study reports much higher functional completion than joint functionality and security success across six coding agents. Research intake: DB-RADAR-CP7-CP7A-002.
Claim-level citations (1)
- supportsSecure Code Benchmark separates working agent code from security-correct code: The cited authors or maintainers report the following bounded finding: A 450-task study reports much higher functional completion than joint functionality and security success across six coding agents.
Original primary document, reported finding and evaluation context; checkpoint DB-RADAR-CP7-CP7A-002