multi-agent-smart-contracts / data /final_evaluation_results.csv
Tales-Cunha
feat: refactor tester agent to PoCo LangGraph architecture and add synthetic benchmark
017c628
Raw
History Blame Contribute Delete
4.36 kB
ID;Time_Sec;Reproducible;Specific;False_Positive_Rejected;A_Infra_Iters;A_Exploit_Iters;A_Final_Error;B_Infra_Iters;B_Exploit_Iters;B_Final_Error;PoC_Code;Patch_Diff;A_Execution_Logs;B_Execution_Logs
001;111;FALSE;FALSE;TRUE;0;0;Max tool calls (30) exceeded.;0;0;Max tool calls (30) exceeded.;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;;;