Meta harness, from Stanford's Omar Khattab (creator of DSPy), treats the harness as an optimization target by using an agentic proposer (Claude Opus 4.6) that reads failed execution traces, diagnoses breakage, and writes complete new harnesses, with evaluators testing each proposal and repeating the loop, consuming 10 million tokens per iteration (400 times more feedback than prior methods) across 82 files read per round.

factualpending

Speaker

Unidentified Speaker — Rethinking AI Agents: The Rise of Harness Engineering [Xxuxg8PcBvc]

Evidence Quote

“Meta harness from Stanford's Omar Khattab, creator of DSPy, treats the harness as an optimization target. DSPy tunes prompts within a fixed pipeline. Meta harness rewrites the pipeline itself, structure, retrieval, memory, orchestration topology. Here's the loop. An agentic proposer, Claude Code with Opus 4.6, reads failed execution traces, diagnoses what broke, and writes a complete new harness. Scores and raw traces accumulate in a growing file system. An evaluator tests each proposal. Repeat. The scale. 10 million tokens per iteration, 400 times more feedback than any prior method. 82 files read per round.”

Source

Rethinking AI Agents: The Rise of Harness Engineering— PY
Created: 8/12/2026, 2:27:33 AM

Your notes

Loading notes...