# ModularRSI: Modular and Generalizable Recursive Harness Self-Improvement

arXiv: https://arxiv.org/abs/2609.14857 · HTML: https://mlsanity.com/paper/2609.14857/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Placed in Self-evolving LLM agents, one of 13 directions in Synthetic data and self-improvement for LLMs.
- Builds on 35 of its 36 references that are in the map.
- First posted on arXiv 14 Sep 2026.

## Placed in

- Synthetic data and self-improvement for LLMs / [Self-evolving LLM agents](https://mlsanity.com/synth/self-evolving-llm-agents/)

## Builds on (most cited first, 30 of 35)

- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [Terminal-Bench: Benchmarking Agents on Hard, Realistic Tasks in Command Line Interfaces](https://mlsanity.com/paper/2601.11868/) (2026)
- [GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning](https://mlsanity.com/paper/2507.19457/) (2025)
- [Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models](https://mlsanity.com/paper/2510.04618/) (2025)
- [Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agents](https://mlsanity.com/paper/2505.22954/) (2025)
- [Meta-Harness: End-to-End Optimization of Model Harnesses](https://mlsanity.com/paper/2603.28052/) (2026)
- [SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks?](https://mlsanity.com/paper/2509.16941/) (2025)
- [Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses](https://mlsanity.com/paper/2604.25850/) (2026)
- [Self-Harness: Harnesses That Improve Themselves](https://mlsanity.com/paper/2606.09498/) (2026)
- [Self-Adapting Language Models](https://mlsanity.com/paper/2506.10943/) (2025)
- [AutoHarness: improving LLM agents by automatically synthesizing a code harness](https://mlsanity.com/paper/2603.03329/) (2026)
- [Endless Terminals: Scaling RL Environments for Terminal Agents](https://mlsanity.com/paper/2601.16443/) (2026)
- [Rethinking the Evaluation of Harness Evolution for Agents](https://mlsanity.com/paper/2607.12227/) (2026)
- [On Data Engineering for Scaling LLM Terminal Capabilities](https://mlsanity.com/paper/2602.21193/) (2026)
- [Large-Scale Terminal Agentic Trajectory Generation from Dockerized Environments](https://mlsanity.com/paper/2602.01244/) (2026)
- [ClawBench: Can AI Agents Complete Everyday Online Tasks?](https://mlsanity.com/paper/2604.08523/) (2026)
- [From Failed Trajectories to Reliable LLM Agents: Diagnosing and Repairing Harness Flaws](https://mlsanity.com/paper/2606.06324/) (2026)
- [HarnessForge: Joint Harness and Policy Evolution for Adaptive Agent Systems](https://mlsanity.com/paper/2606.01779/) (2026)
- [Terminal-World: Scaling Terminal-Agent Environments via Agent Skills](https://mlsanity.com/paper/2605.20876/) (2026)
- [Evolving Agents in the Dark: Retrospective Harness Optimization via Self-Preference](https://mlsanity.com/paper/2606.05922/) (2026)
- [A Self-Evolving Framework for Efficient Terminal Agents via Observational Context Compression](https://mlsanity.com/paper/2604.19572/) (2026)
- [HarnessBank: Semantic Gene-Bank Search with Gated Verification for Agent-Harness Self-Evolution](https://mlsanity.com/paper/2607.13683/) (2026)
- [HarnessDev: Can LLMs Create and Evolve Their Own Agent Harness?](https://mlsanity.com/paper/2609.01437/) (2026)
- [OpenSkill: Open-World Self-Evolution for LLM Agents](https://mlsanity.com/paper/2606.06741/) (2026)
- [Recursive Self-Evolving Agents via Held-Out Selection](https://mlsanity.com/paper/2606.28374/) (2026)
- [Watch Before You Answer: Learning from Visually Grounded Post-Training](https://mlsanity.com/paper/2604.05117/) (2026)
- [HiRAS: A Hierarchical Multi-Agent Framework for Paper-to-Code Generation and Execution](https://mlsanity.com/paper/2604.17745/) (2026)
- [RewardHarness: Self-Evolving Agentic Post-Training](https://mlsanity.com/paper/2605.08703/) (2026)
- [Learning from the Self-future: On-policy Self-distillation for dLLMs](https://mlsanity.com/paper/2606.18195/) (2026)
- [S3Gym: Can LLMs Turn Self-Testing and Self-Judging into Self-Improvement?](https://mlsanity.com/paper/2608.31100/) (2026)

## Built on by (most cited first, 9 of 9)

- [VERSE: Verified Self-Evolving Optimizer for Agent Harnesses](https://mlsanity.com/paper/2610.02616/) (2026)
- [Scaling Trajectories for Complex Tasks through Recursive Self-Rewrite](https://mlsanity.com/paper/2610.02826/) (2026)
- [MemCodex: Self-Programming Hierarchical Memory for Language Agents](https://mlsanity.com/paper/2609.39765/) (2026)
- [MMPostTrainBench: Benchmarking Autonomous Research for Multimodal Post-Training](https://mlsanity.com/paper/2610.05398/) (2026)
- [What Does a Harness Repair? A Preregistered Study of Visibility, Baseline Adequacy and Evaluation Defects](https://mlsanity.com/paper/2610.05533/) (2026)
- [VERA: Scaling Verifiable Environments for Agentic co-Evolution](https://mlsanity.com/paper/2610.05923/) (2026)
- [DuplexAgent-RSI: Recursive Harness Improvement for Full-Duplex Voice Agent Collaboration](https://mlsanity.com/paper/2610.11299/) (2026)
- [From Existing Methods to New Research Ideas: Two Perspectives on Recursive Self-Improvement](https://mlsanity.com/paper/2610.14625/) (2026)
- [Self-Evolving Agent Harness](https://mlsanity.com/paper/2610.18813/) (2026)

## Abstract (arXiv)

Recent work extends recursive self-improvement (RSI) to agent harnesses for long-horizon coding and terminal tasks, enabling agents to improve execution mechanisms from experience. However, generalizable harness RSI remains challenging. First, evolving harnesses on evaluation benchmarks or their subsets makes it difficult to distinguish reusable improvements from benchmark-specific adaptation. Second, single-trajectory updates can conflate systematic harness deficiencies with instance-specific reasoning and solution details, producing modifications that transfer poorly to unseen tasks. Third, localizing recurring behavioral deficiencies within monolithic harnesses is difficult, while whole-harness optimization can entangle unrelated mechanisms and complicate attribution and validation. We propose ModularRSI, a benchmark-disjoint, contrastive, and modular framework for generalizable harness evolution. ModularRSI contrasts successful and failed trajectories for the same task and aggregates evidence across tasks to identify recurring behavioral deficiencies. It decomposes the evolvable harness into five functional modules: Agent Loop, Tool Use, Observation Management, Context Management, and Task Completion Detection. Each module evolves independently within a restricted modification scope, followed by an integration stage that combines the evolved modules into a unified harness and resolves potential conflicts. To support benchmark-disjoint evolution, we curate 2,000 executable evolution tasks from external sources that are disjoint from downstream evaluation benchmarks. Experiments on TB2.0 and SWE-Bench Verified show consistent improvements on unseen in-domain and cross-domain tasks, with the evolved harness also transferring across different foundation models.