List your company

Red-teaming tool-calling agents on MCP Gmail 1.0 (test)

0.844RRZero-Shot (ZS)
0.0640.26650.4690.6715Mar 21, 2026

Evaluation results

MethodLinks
Zero-Shot (ZS)Mode=Zero-Shot2026.030.8440
Multi-Turn (MT)Mode=Multi-Turn2026.030.6560.031
Iterative Refinement (IR)Mode=Iterative Refinement2026.030.4530.156
Standard Evolution (SE)Mode=Evolutionary Search2026.030.2030.109
T-MAPBackbone=DeepSeek-V3.2, Target Model=GPT-5-mini2026.030.0940.469