감사 이력
agentic-structure - 6 감사
감사 버전 6
최신 낮은 위험Jun 28, 2026, 01:12 PM
Static analysis reported many critical and high patterns, but review found them to be documentation false positives, mostly Markdown backticks, security terminology, and examples that warn against unsafe practices. No prompt injection, credential access, malware behavior, or command execution intent was found. The remaining low risk is that the knowledge protocol describes creating reference files and using web fetches for documentation.
낮은 위험 문제 (4)
위험 요인
📁 파일 시스템 접근 (4)
감사 버전 5
안전Jan 16, 2026, 05:20 PM
Pure documentation skill containing only markdown guidelines. No executable code, no network operations, no command execution. The skill specifies allowed-tools: Read, Grep, Glob only. All 142 static findings are FALSE POSITIVES caused by the scanner detecting security terminology in documentation (e.g., 'Do not use MD5' flagged as 'weak crypto').
위험 요인
⚙️ 외부 명령어 (36)
감사 버전 4
안전Jan 16, 2026, 05:20 PM
Pure documentation skill containing only markdown guidelines. No executable code, no network operations, no command execution. The skill specifies allowed-tools: Read, Grep, Glob only. All 142 static findings are FALSE POSITIVES caused by the scanner detecting security terminology in documentation (e.g., 'Do not use MD5' flagged as 'weak crypto').
위험 요인
⚙️ 외부 명령어 (36)
감사 버전 3
안전Jan 10, 2026, 10:55 AM
Pure prompt-based skill containing only markdown documentation files. No executable code, network operations, filesystem access, or command execution. The skill provides development guidelines for AI coding assistants.
감사 버전 2
안전Jan 10, 2026, 10:55 AM
Pure prompt-based skill containing only markdown documentation files. No executable code, network operations, filesystem access, or command execution. The skill provides development guidelines for AI coding assistants.
감사 버전 1
안전Jan 10, 2026, 10:55 AM
Pure prompt-based skill containing only markdown documentation files. No executable code, network operations, filesystem access, or command execution. The skill provides development guidelines for AI coding assistants.