Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Security Robustness on MCPSafetybench
Loading...
5
ASR (w/o)
MCPShield
3.8
11.9
20
28.1
Feb 15, 2026
ASR (w/o)
ASR (Ave)
ASR (Min)
ASR (Max)
Updated 1mo ago
Evaluation Results
Method
Method
Links
ASR (w/o)
ASR (Ave)
ASR (Min)
ASR (Max)
MCPShield
Backbone=Gemini3-Flash
2026.02
5
100
100
100
MCPShield
Backbone=Kimi-K2
2026.02
5
84.44
83.33
88.89
MCPShield
Backbone=Qwen3 235B
2026.02
5
100
100
100
MCPShield
Backbone=Deepseek V3.2
2026.02
10
100
100
100
MCPShield
Backbone=GPT4o-mini
2026.02
15
94.44
94.44
94.44
MCPShield
Backbone=Minimax-M2
2026.02
35
84.04
66.67
100
Feedback
Search any
task
Search any
task