# SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness > Source: > Published: 2026-09-18 02:00:16+00:00 As coding agents move from supervised code completion to unattended, around-the-clock exploration, their work expands from isolated predictions into long trajectories of reasoning, tool use, and feedback. Token efficiency therefore becomes important for scaling recursive self-improvement. We take an