# COMPASS: Enhancing Agent Long-Horizon Reasoning with Evolving Context

arXiv: https://arxiv.org/abs/2510.08790 · HTML: https://mlsanity.com/paper/2510.08790/

## Key facts

- Cited by 20 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning.
- Builds on 24 of its 26 references that are in the map.
- First posted on arXiv 9 Oct 2025.

## Placed in

- LLM reasoning / [Reasoning for agents, planning and code](https://mlsanity.com/lr/reasoning-for-agents-planning-and-code/)
- LLM agents and tool use / [General LLM agent frameworks and benchmarks](https://mlsanity.com/agents/general-llm-agent-frameworks-and-benchmarks/)

## Builds on (most cited first, 24 of 24)

- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023)
- [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023)
- [Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learning](https://mlsanity.com/paper/2503.09516/) (2025)
- [CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Society](https://mlsanity.com/paper/2303.17760/) (2023)
- [GAIA: a benchmark for General AI Assistants](https://mlsanity.com/paper/2311.12983/) (2023)
- [Large Language Models Cannot Self-Correct Reasoning Yet](https://mlsanity.com/paper/2310.01798/) (2023)
- [Encouraging Divergent Thinking in Large Language Models through Multi-Agent Debate](https://mlsanity.com/paper/2305.19118/) (2023)
- [BrowseComp: A Simple Yet Challenging Benchmark for Browsing Agents](https://mlsanity.com/paper/2504.12516/) (2025)
- [Multi-Agent Collaboration Mechanisms: A Survey of LLMs](https://mlsanity.com/paper/2501.06322/) (2025)
- [Understanding the planning of LLM agents: A survey](https://mlsanity.com/paper/2402.02716/) (2024)
- [ReTool: Reinforcement Learning for Strategic Tool Use in LLMs](https://mlsanity.com/paper/2504.11536/) (2025)
- [LongRoPE: Extending LLM Context Window Beyond 2 Million Tokens](https://mlsanity.com/paper/2402.13753/) (2024)
- [Prompt Cache: Modular Attention Reuse for Low-Latency Inference](https://mlsanity.com/paper/2311.04934/) (2023)
- [AdaPlanner: Adaptive Planning from Feedback with Language Models](https://mlsanity.com/paper/2305.16653/) (2023)
- [Plan-and-Act: Improving Planning of Agents for Long-Horizon Tasks](https://mlsanity.com/paper/2503.09572/) (2025)
- [Deep Research Agents: A Systematic Examination And Roadmap](https://mlsanity.com/paper/2506.18096/) (2025)
- [When Thinking Fails: The Pitfalls of Reasoning for Instruction-Following in LLMs](https://mlsanity.com/paper/2505.11423/) (2025)
- [Human-In-the-Loop Software Development Agents](https://mlsanity.com/paper/2411.12924/) (2024)
- [Robotouille: An Asynchronous Planning Benchmark for LLM Agents](https://mlsanity.com/paper/2502.05227/) (2025)
- [LLM-based Automated Grading with Human-in-the-Loop](https://mlsanity.com/paper/2504.05239/) (2025)
- [PLAN-TUNING: Post-Training Language Models to Learn Step-by-Step Planning for Complex Problem Solving](https://mlsanity.com/paper/2507.07495/) (2025)

## Built on by (most cited first, 20 of 20)

- [SWE-Pruner: Self-Adaptive Context Pruning for Coding Agents](https://mlsanity.com/paper/2601.16746/) (2026)
- [A Self-Evolving Framework for Efficient Terminal Agents via Observational Context Compression](https://mlsanity.com/paper/2604.19572/) (2026)
- [The Evolution of Tool Use in LLM Agents: From Single-Tool Call to Multi-Tool Orchestration](https://mlsanity.com/paper/2603.22862/) (2026)
- [FoldAct: Efficient and Stable Context Folding for Long-Horizon Search Agents](https://mlsanity.com/paper/2512.22733/) (2025)
- [CoACT: Action-Preserving Observation Compression for Coding Agents](https://mlsanity.com/paper/2607.02911/) (2026)
- [ConvexBench: Can LLMs Recognize Convex Functions?](https://mlsanity.com/paper/2602.01075/) (2026)
- [Ecdysis: Efficient and Effective Training of Runtime Harnesses for LLM Agents](https://mlsanity.com/paper/2609.11677/) (2026)
- [Context Pruning for Coding Agents via Multi-Rubric Latent Reasoning](https://mlsanity.com/paper/2605.15315/) (2026)
- [SWE-Pruner Pro: The Coder LLM Already Knows What to Prune](https://mlsanity.com/paper/2607.18213/) (2026)
- [The Convergence of Schema-Guided Dialogue Systems and the Model Context Protocol](https://mlsanity.com/paper/2602.18764/) (2026)
- [ReflectCAP: Detailed Image Captioning with Reflective Memory](https://mlsanity.com/paper/2604.12357/) (2026)
- [Remember Your Trace: Memory-Guided Long-Horizon Agentic Framework for Consistent and Hierarchical Repository-Level Code Documentation](https://mlsanity.com/paper/2605.14563/) (2026)
- [TrajWiki: Source-Grounded Memory Trajectories for Long-Horizon Dialogue Agents](https://mlsanity.com/paper/2608.00967/) (2026)
- [ThinkReset: Learnable Intermediate Interface Construction for Bounded-Context Long-Horizon Reasoning](https://mlsanity.com/paper/2607.28642/) (2026)
- [DTOC: Dynamic Tool Output Compression for Adaptive Context Management in AI Agents](https://mlsanity.com/paper/2609.26121/) (2026)
- [Adaptive Consistency Graph for Long-Horizon Agents](https://mlsanity.com/paper/2609.32754/) (2026)
- [Beyond Skill Evolution: Self-Evolving Context Management Policies for Long-Horizon Agent Harnesses](https://mlsanity.com/paper/2609.34649/) (2026)
- [Continuous Context Management](https://mlsanity.com/paper/2609.35540/) (2026)
- [Beyond Memory: Harnessing Long-Horizon Agents with Explicit Belief States](https://mlsanity.com/paper/2610.01415/) (2026)
- [Beyond Correctness: Resolving Underspecification in Agentic Text-to-SQL](https://mlsanity.com/paper/2610.02739/) (2026)

## Abstract (arXiv)

Long-horizon tasks that require sustained reasoning and multiple tool interactions remain challenging for LLM agents: small errors compound across steps, and even state-of-the-art models often hallucinate or lose coherence. We identify context management as the central bottleneck -- extended histories cause agents to overlook critical evidence or become distracted by irrelevant information, thus failing to replan or reflect from previous mistakes. To address this, we propose COMPASS (Context-Organized Multi-Agent Planning and Strategy System), a lightweight hierarchical framework that separates tactical execution, strategic oversight, and context organization into three specialized components: (1) a Main Agent that performs reasoning and tool use, (2) a Meta-Thinker that monitors progress and issues strategic interventions, and (3) a Context Manager that maintains concise, relevant progress briefs for different reasoning stages. Across three challenging benchmarks -- GAIA, BrowseComp, and Humanity's Last Exam -- COMPASS improves accuracy by up to 20% relative to both single- and multi-agent baselines. We further introduce a test-time scaling extension that elevates performance to match established DeepResearch agents, and a post-training pipeline that delegates context management to smaller models for enhanced efficiency.