Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Malicious Skill Detection on 100-skill dataset (OpenClaw)
Loading...
98
TPR
SkillScanner (LLM)
-3.92
22.54
49
75.46
Jun 10, 2026
TPR
FPR
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
TPR
FPR
Accuracy
SkillScanner (LLM)
variant=LLM-judgment
2026.06
98
90
54
RSA
2026.06
88
8
90
SkillScanner
variant=default
2026.06
70
12
77
Aguara
2026.06
64
14
75
Semgrep (custom)
ruleset=custom
2026.06
34
0
67
ClawScan
2026.06
2
6
48
Semgrep (generic)
ruleset=generic
2026.06
0
2
49
Feedback
Search any
task
Search any
task