205 B
205 B
promptlooper
Universal LLM pipeline tuning workbench — systematically optimize prompts, models, and inference parameters through cached experiments, pluggable scoring, and agent-driven sweeps via MCP.