The Scaffold Effect in Coding Agents: Harness Choice as a Hidden Variable in Coding-Agent Evaluation

• URL: http://arxiv.org/abs/2607.22585v1
• Date: Mon, 08 Jun 2026 20:27:06 GMT
• Title: The Scaffold Effect in Coding Agents: Harness Choice as a Hidden Variable in Coding-Agent Evaluation
• Abstract summary: We evaluate Qwen 3.6 Plus and MiniMax M2.5 across three open-source harnesses.<n> Harness choice induces up to a 40x difference in tokens per solved task.<n>We recommend selecting harness-model pairs by pass rate under token/latency budgets.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Public leaderboards for coding agents typically rank systems by model name and pass rate, while the surrounding harness (the scaffold that issues tools, manages context, and decides when to stop) is often under-specified. Model-to-model comparison is valid when the harness is fixed; when it varies, performance and efficiency conflate model and scaffold effects. We evaluate Qwen 3.6 Plus and MiniMax M2.5 across three open-source harnesses (Goose, OpenCode, OpenHands-SDK) on a stratified 50-task subset of Terminal-Bench Pro. Harness choice induces up to a 40x difference in tokens per solved task, while paired within-model pass-rate differences remain 0-8 percentage points (95% paired-task bootstrap CIs include zero except for the largest gap). Failure fingerprints replicate across models (REASON for Goose, VERIFY/MAX_TURNS for OpenHands-SDK, idle-loop/TIME for OpenCode), indicating harness-level biases that are largely model-independent. For human-centered coding-agent evaluation, model name alone is an incomplete comparison unit: harness-model pairs determine real-world cost, latency, and oversight burden; no-action turns are a per-task wait tax, not just a token tax. We therefore recommend selecting harness-model pairs by pass rate under token/latency budgets, and reporting token usage, latency, and full harness specifications alongside any model comparison. We release anonymized configs, raw trial logs, aggregated snapshots, and analysis scripts.

Related papers

• HarnessDev: Can LLMs Create and Evolve Their Own Agent Harness? [61.02699867449589]
  We introduce HarnessDev, a benchmark that shifts the unit of evaluation from task outputs to run infrastructure.<n>In Creation, the agent starts from a minimal seed and a small number of cases, then builds a complete execution system.<n>In Evolution, it starts from its own created harness and iteratively revises it using downstream execution feedback, with the goal of improving benchmark performance.
  arXiv  Detail & Related papers  (2026-09-01T15:45:33Z)
• Same Model, Different Harness: Different Coding-Agent Results [0.0]
  We compare two configurations of the same harness on three coding benchmarks.<n>The control supplies the full conversation in time order, while the treatment keeps the same record but mechanically shortens older tool results.<n>Under tight context, the treatment raises mean per-task fail-to-pass fraction (F2PF) in all three pressure comparisons.
  arXiv  Detail & Related papers  (2026-08-26T11:55:44Z)
• Agent Behavioral Contracts II: Certifying Compositional Reliability Without Assuming Independence [0.0]
  Two instances of one model, in a two-agent handoff, co-fail on 90.0% of the missions on which either fails.<n>Substituting a different model reduces the association in six of six contrasts.
  arXiv  Detail & Related papers  (2026-08-13T07:25:05Z)
• Surrogate Fidelity: When Can Open LLMs Explain Closed Ones? [6.786301369955054]
  We evaluate surrogate fidelity at the prediction, attribution, and representation levels.<n>We document an access-validity inversion: white-box signals like attention patterns and magnitudes are highly stable across models but only weakly predictive of causal attributions.
  arXiv  Detail & Related papers  (2026-06-30T17:43:44Z)
• Cluster, Route, Escalate: Cascaded Framework for Cost-Aware LLM Serving [69.5155152169752]
  We propose a two-stage cascaded solution for large language models (LLMs)<n>Stage 1 clusters incoming queries and assigns each cluster to its most cost-effective model.<n>Stage 2 adds a quality estimation (QE) cascade; when an output from Stage 1 is judged low-quality, the query is escalated to a stronger model.<n>On the test datasets, the cascaded system retains 97-99% of the strongest model's accuracy while reducing Time Per Output Token (TPOT)<n>It requires only task-correctness labels and adapts to changes in the model pool without manual reconfiguration.
  arXiv  Detail & Related papers  (2026-06-25T18:29:24Z)
• Dissecting model behavior through agent trajectories [15.811597127775812]
  We formalize the intent-execution' gap: the mismatch between what the model intends and what the harness executes, and vice versa.<n>To illustrate the impact of this harness-model alignment, we develop a simple and customizable harness called Simple Strands Agent' (SSA)<n>We make two contributions: (i) we reproduce or improve on the pass@1 performance reported by diverse model-provider families on popular agentic benchmarks, and (ii) building on an analysis of 138k trajectories generated by SSA.
  arXiv  Detail & Related papers  (2026-06-16T03:17:03Z)
• Catch Your Breath: Adaptive Computation for Self-Paced Sequence Production [55.76222360698305]
  We explore a class of supervised training objectives that allow a language model to dynamically and autonomously scale the number of compute steps used for each input token.<n>For any token, the model can request additional compute steps by emitting a don't know> output.<n>We find that the CYB model requests additional steps when doing so improves accuracy, and the model adapts its processing time to token-level complexity and context.
  arXiv  Detail & Related papers  (2025-10-13T21:07:05Z)
• Every Step Counts: Decoding Trajectories as Authorship Fingerprints of dLLMs [63.82840470917859]
  We show that the decoding mechanism of dLLMs can be used as a powerful tool for model attribution.<n>We propose a novel information extraction scheme called the Directed Decoding Map (DDM), which captures structural relationships between decoding steps and better reveals model-specific behaviors.
  arXiv  Detail & Related papers  (2025-10-02T06:25:10Z)
• Ranked from Within: Ranking Large Multimodal Models Without Labels [73.96543593298426]
  We show that uncertainty scores derived from softmax distributions provide a robust basis for ranking models across various tasks.<n>This facilitates the ranking of LMMs on unlabeled data, providing a practical approach for selecting models for diverse target domains without requiring manual annotation.
  arXiv  Detail & Related papers  (2024-12-09T13:05:43Z)
• When Liebig's Barrel Meets Facial Landmark Detection: A Practical Model [87.25037167380522]
  We propose a model that is accurate, robust, efficient, generalizable, and end-to-end trainable. In order to achieve a better accuracy, we propose two lightweight modules. DQInit dynamically initializes the queries of decoder from the inputs, enabling the model to achieve as good accuracy as the ones with multiple decoder layers. QAMem is designed to enhance the discriminative ability of queries on low-resolution feature maps by assigning separate memory values to each query rather than a shared one.
  arXiv  Detail & Related papers  (2021-05-27T13:51:42Z)
• AvgOut: A Simple Output-Probability Measure to Eliminate Dull Responses [97.50616524350123]
  We build dialogue models that are dynamically aware of what utterances or tokens are dull without any feature-engineering. The first model, MinAvgOut, directly maximizes the diversity score through the output distributions of each batch. The second model, Label Fine-Tuning (LFT), prepends to the source sequence a label continuously scaled by the diversity score to control the diversity level. The third model, RL, adopts Reinforcement Learning and treats the diversity score as a reward signal.
  arXiv  Detail & Related papers  (2020-01-15T18:32:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.