elder-plinius/T3MP3ST

This request for 'benchmarks per model / harness' highlights a critical market demand for quantifiable performance metrics in autonomous red teaming platforms. Users require empirical data to assess the efficacy and reliability of different AI models and configurations within T3MP3ST. The current 'sometimes' effectiveness of prompts underscores the inconsistency and lack of transparency. For a B2B offensive security solution, performance benchmarks are not merely a feature; they are a prerequisite for trust and adoption. Customers need to justify investments with clear ROI, which necessitates understanding how different models perform against various threat scenarios. Providing these benchmarks would significantly enhance the platform's credibility and facilitate informed decision-making for security teams.