# The Working Set of a Coding Agent: Coherence Debt in Repository-Scale Tasks arXiv: https://arxiv.org/abs/2608.16630 ยท HTML: https://mlsanity.com/paper/2608.16630/ ## Key facts - Cited by 1 papers in the map (data through 14 Oct 2026). - Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use. - Builds on 23 of its 36 references that are in the map. - First posted on arXiv 17 Aug 2026. ## Placed in - LLM agents and tool use / [Software engineering coding agents](https://mlsanity.com/agents/software-engineering-coding-agents/) ## Builds on (most cited first, 23 of 23) - [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022) - [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023) - [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023) - [MemGPT: Towards LLMs as Operating Systems](https://mlsanity.com/paper/2310.08560/) (2023) - [RULER: What's the Real Context Size of Your Long-Context Language Models?](https://mlsanity.com/paper/2404.06654/) (2024) - [OpenHands: An Open Platform for AI Software Developers as Generalist Agents](https://mlsanity.com/paper/2407.16741/) (2024) - [RepoGraph: Enhancing AI Software Engineering with Repository-level Code Graph](https://mlsanity.com/paper/2410.14684/) (2024) - [Understanding Software Engineering Agents: A Study of Thought-Action-Result Trajectories](https://mlsanity.com/paper/2506.18824/) (2025) - [Understanding Code Agent Behaviour: An Empirical Study of Success and Failure Trajectories](https://mlsanity.com/paper/2511.00197/) (2025) - [Evaluating AGENTS.md: Are Repository-Level Context Files Helpful for Coding Agents?](https://mlsanity.com/paper/2602.11988/) (2026) - [AgentLens: Revealing The Lucky Pass Problem in SWE-Agent Evaluation](https://mlsanity.com/paper/2605.12925/) (2026) - [Failure as a Process: An Anatomy of CLI Coding Agent Trajectories](https://mlsanity.com/paper/2607.09510/) (2026) - [TrajAudit: Automated Failure Diagnosis for Agentic Coding Systems](https://mlsanity.com/paper/2605.26563/) (2026) - [The Missing Memory Hierarchy: Demand Paging for LLM Context Windows](https://mlsanity.com/paper/2603.09023/) (2026) - [CORE-Bench: A Comprehensive Benchmark for Code Retrieval in the Era of Agentic Coding](https://mlsanity.com/paper/2606.11864/) (2026) - [SE-Bench: Benchmarking Self-Evolution with Knowledge Internalization](https://mlsanity.com/paper/2602.04811/) (2026) - [Coherence Collapse: Diagnosing Why Code Agents Fail After Reaching the Right Code](https://mlsanity.com/paper/2603.24631/) (2026) - [When Parallelism Pays Off: Cohesion-Aware Task Partitioning for Multi-Agent Coding](https://mlsanity.com/paper/2606.00953/) (2026) - [ClawVM: Harness-Managed Virtual Memory for Stateful Tool-Using LLM Agents](https://mlsanity.com/paper/2604.10352/) (2026) - [SWE-Edit: Rethinking Code Editing for Efficient SWE-Agent](https://mlsanity.com/paper/2604.26102/) (2026) - [Strained Coherence: A Pre-Failure Signal in Coding Agent Execution Trajectories](https://mlsanity.com/paper/2606.07889/) (2026) - [Token Coherence: Adapting MESI Cache Protocols to Minimize Synchronization Overhead in Multi-Agent LLM Systems](https://mlsanity.com/paper/2603.15183/) (2026) - [How Task Structure Limits Multi-Agent Success: An Information-Theoretic Analysis](https://mlsanity.com/paper/2606.13733/) (2026) ## Built on by (most cited first, 1 of 1) - [StateTape: Action-Conditioned Evidence Lifecycle Modeling for Long-Horizon Coding Agents](https://mlsanity.com/paper/2609.36319/) (2026) ## Abstract (arXiv) Repository-scale coding requires an agent to keep tests, imports, configuration, and migration rules consistent within a bounded context window. We model this as reconstructing a coupled-fact graph: at each edit, a required fact comes from recent context or parametric memory, and the facts covered by neither form coherence debt. We supply and withhold each channel and inject faults across seven models and five harnesses. As expected, no model completes a task on an unseen API with both channels empty, and putting the facts in the prompt restores success. When a rename defeats what models memorized about a real library, all seven fail in the same place, passing and missing the same tests. Availability decides the outcome and distance does not: withholding a fact costs exactly the work it supports, and a supplied fact works as well far from the edit as next to it. Harnesses pay unequal prices for it: configurations that all pass every test differ more than tenfold in tokens consumed because they rebuild the same content at different rates, and spending more recovers nothing when facts are withheld. A missing fact produces wrong work rather than absent work: an agent asked to act acts, fabricating the file or guessing the value, so instruments built on reads look for a hole already filled. How often it says it is blocked instead is a property of the model, from every trial to none. Availability does not settle every edit: where standard and code disagree, agents follow the standard even when it prescribes the worse code, so a stale convention file costs more than no file. Because parametric memory substitutes for reading, on SWE-bench, where models likely know the repositories, reads no longer predict success. Harnesses should keep the facts an edit depends on available when the agent writes, and check that availability against what the agent produces rather than what it reads.