Stop Comparing LLM Agents Without Disclosing the Harness
For researchers and practitioners evaluating LLM agents, this work highlights a critical confound in current benchmarks, proposing a disclosure standard to improve evaluation validity.
This paper argues that for long-horizon agent tasks with comparable frontier models, the execution harness (infrastructure for context, tools, orchestration) often determines performance more than the model itself, and current evaluations misattribute harness gains to model improvements. Controlled variance decomposition shows harness-induced variance can exceed model-induced variance, including ranking reversals.
This position paper argues that, for long-horizon tasks evaluated across models with comparable frontier capability, the agent execution harness, namely the infrastructure layer that governs context construction, tool interaction, orchestration, and verification around a language model, is often a stronger determinant of agent performance than the model it wraps. We formalize and defend the Binding Constraint Thesis: in this regime, performance variance is governed more by harness configuration than by model choice, and current evaluation protocols therefore systematically misattribute harness-level gains to model improvements. We support this thesis along three lines. First, a control-theoretic formalization treats the harness as the controller of a closed-loop dynamical system and the LLM as the stochastic policy it governs, which explains why small harness changes can produce performance shifts that exceed those obtained by substituting one model for another. Second, published benchmarks, industry deployments, and a controlled variance decomposition show that harness-induced variance can substantially exceed model-induced variance, including cases of model ranking reversal. Third, we propose a harness-aware evaluation framework with a disclosure standard and a variance decomposition protocol. Until harness specifications are disclosed, leaderboard comparisons for long-horizon agents should be treated as incomplete and potentially misleading.