When the Specification Emerges: Benchmarking Faithfulness Loss in Long-Horizon Coding Agents

TL;DR

This paper introduces SLUMP, a benchmark for evaluating how well long-horizon coding agents maintain faithfulness when the task specification emerges gradually through interaction, highlighting challenges and potential mitigation strategies.

Contribution

It presents a novel benchmark and evaluation methodology for faithfulness loss in emergent specification settings, along with a case study of a mitigation approach called ProjectGuard.

Findings

Structural integration degrades under emergent specification.

Semantic faithfulness loss varies between models, being substantial on Claude Code.

ProjectGuard significantly recovers faithfulness and reduces failures.

Abstract

Current coding-agent benchmarks usually pro- vide the full task specification upfront. Real research coding often does not: the intended system is progressively disclosed through in- teraction, requiring the agent to track durable design commitments across a long session. We introduce a benchmark for this setting and study faithfulne Ss Loss U nder eM ergent s Pecification (SLUMP), defined as the reduc- tion in final implementation faithfulness un- der emergent specification relative to a single- shot specification control. The benchmark con- tains 20 recent ML papers (10 ICML 2025, 10 NeurIPS 2025), 371 atomic verifiable compo- nents, and interaction scripts of approximately 60 coding requests that progressively disclose the target design without revealing the paper itself. Final repositories are scored with a five-level component-faithfulness rubric and accompanied by an exposure audit to verify that scored components are recoverable from the visible interaction. Evaluated on Claude Code and Codex, the single-shot specification control achieves higher overall implementation fidelity on 16/20 and 14/20 papers, respectively. Structural integration degrades under emergent specification on both platforms, while seman- tic faithfulness loss is substantial on Claude Code and small on Codex. As a mitigation case study, we introduce ProjectGuard, an exter- nal project-state layer for specification tracking. On Claude Code, ProjectGuard recovers 90% of the faithfulness gap, increases fully faith- ful components from 118 to 181, and reduces severe failures from 72 to 49. These results identify specification tracking as a distinct eval- uation target for long-horizon coding agents.