List your company

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

100RRZero-Shot (ZS)
10.66433.85757.0580.243Mar 21, 2026

Evaluation results

MethodLinks
Zero-Shot (ZS)Mode=Zero-Shot2026.031000
Multi-Turn (MT)Mode=Multi-Turn2026.0373.41.6
Iterative Refinement (IR)Mode=Iterative Refinement2026.0370.33.1
Standard Evolution (SE)Mode=Evolutionary Search2026.0317.248.4
T-MAPBackbone=DeepSeek-V3.2, Target Model=GPT-5-mini2026.0314.156.2