# GRASP: Gated Regression-Aware Skill Proposer for Self-Improving LLM Agents

arXiv: https://arxiv.org/abs/2605.29668 · HTML: https://mlsanity.com/paper/2605.29668/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Placed in Agent skill learning and evolution, one of 16 directions in LLM agents and tool use.
- Builds on 25 of its 33 references that are in the map.
- First posted on arXiv 28 May 2026.

## Placed in

- LLM agents and tool use / [Agent skill learning and evolution](https://mlsanity.com/agents/agent-skill-learning-and-evolution/)
- Synthetic data and self-improvement for LLMs / [Self-evolving LLM agents](https://mlsanity.com/synth/self-evolving-llm-agents/)

## Builds on (most cited first, 25 of 25)

- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023)
- [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)
- [WebArena: A Realistic Web Environment for Building Autonomous Agents](https://mlsanity.com/paper/2307.13854/) (2023)
- [MemGPT: Towards LLMs as Operating Systems](https://mlsanity.com/paper/2310.08560/) (2023)
- [AgentBench: Evaluating LLMs as Agents](https://mlsanity.com/paper/2308.03688/) (2023)
- [WebShop: Towards Scalable Real-World Web Interaction with Grounded Language Agents](https://mlsanity.com/paper/2207.01206/) (2022)
- [Mind2Web: Towards a Generalist Agent for the Web](https://mlsanity.com/paper/2306.06070/) (2023)
- [ALFWorld: Aligning Text and Embodied Environments for Interactive Learning](https://mlsanity.com/paper/2010.03768/) (2020)
- [A-MEM: Agentic Memory for LLM Agents](https://mlsanity.com/paper/2502.12110/) (2025)
- [Large Language Models Are Human-Level Prompt Engineers](https://mlsanity.com/paper/2211.01910/) (2022)
- [ExpeL: LLM Agents Are Experiential Learners](https://mlsanity.com/paper/2308.10144/) (2023)
- [Large Language Models as Optimizers](https://mlsanity.com/paper/2309.03409/) (2023)
- [Large Language Models Can Be Easily Distracted by Irrelevant Context](https://mlsanity.com/paper/2302.00093/) (2023)
- [DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines](https://mlsanity.com/paper/2310.03714/) (2023)
- [When Can LLMs Actually Correct Their Own Mistakes? A Critical Survey of Self-Correction of LLMs](https://mlsanity.com/paper/2406.01297/) (2024)
- [Where LLM Agents Fail and How They can Learn From Failures](https://mlsanity.com/paper/2509.25370/) (2025)
- [SkillX: Automatically Constructing Skill Knowledge Bases for Agents](https://mlsanity.com/paper/2604.04804/) (2026)
- [FHIR-AgentBench: Benchmarking LLM Agents for Realistic Interoperable EHR Question Answering](https://mlsanity.com/paper/2509.19319/) (2025)
- [Sample-Efficient Online Learning in LM Agents via Hindsight Trajectory Rewriting](https://mlsanity.com/paper/2510.10304/) (2025)
- [Experiential Reflective Learning for Self-Improving LLM Agents](https://mlsanity.com/paper/2603.24639/) (2026)
- [How Do LLMs Fail In Agentic Scenarios? A Qualitative Analysis of Success and Failure Scenarios of Various LLMs in Agentic Simulations](https://mlsanity.com/paper/2512.07497/) (2025)
- [Evo-MedAgent: Beyond One-Shot Diagnosis with Agents That Remember, Reflect, and Improve](https://mlsanity.com/paper/2604.14475/) (2026)

## Built on by (most cited first, 11 of 11)

- [The Regression Tax: Decomposing Why Skills Help and Hurt LLM Agents](https://mlsanity.com/paper/2607.22520/) (2026)
- [Reconstruct, Practice, Go Real: Guided Self-Improvement for Embodied Agents](https://mlsanity.com/paper/2610.02204/) (2026)
- [ADIAS: Automated Design of Interactive Agentic Systems](https://mlsanity.com/paper/2608.06410/) (2026)
- [Agents in the Large: Perception-Centered Architecture for Persistent Agents](https://mlsanity.com/paper/2608.30478/) (2026)
- [More Programs or More Rolls? Separating Coverage from Specialization in LLM Harnesses](https://mlsanity.com/paper/2609.35873/) (2026)
- [Regimes: An Auditable, Held-Out-Gated Improvement Loop Demonstrated on LongMemEval with ActiveGraph](https://mlsanity.com/paper/2606.10241/) (2026)
- [FRAMES: Guarded and Dual-Objective Skill Evolution for Agents in Policy-Governed Enterprise Workflows](https://mlsanity.com/paper/2608.01772/) (2026)
- [Local Edits, Global Ripples: Replay-Informed Policy Adaptation for Workflow Synthesis](https://mlsanity.com/paper/2609.12127/) (2026)
- [StraTune: Adaptive Selection of Revision Operators for Self-Evolving LLM Skills](https://mlsanity.com/paper/2609.32886/) (2026)
- [Retrieval-Augmented Skill Optimization via Cross-Harness Adaptation](https://mlsanity.com/paper/2609.38024/) (2026)
- [Training Numerical Intelligence via Auto-Diagnosis and Skill Discovery](https://mlsanity.com/paper/2610.03872/) (2026)

## Abstract (arXiv)

LLM agents acting in structured environments fail in operational rather than conversational ways, and reliability depends on procedural knowledge of the environment. Prior self-improvement methods accumulate natural-language guidance without checking that each new item preserves previously correct behavior, so a note that fixes one trajectory can silently regress another. We introduce GRASP (Gated Regression-Aware Skill Proposer), which treats agent improvement as a sequence of edits to a bounded skill library, admitting each candidate only if it produces a net improvement on a balanced held-out probe under a hard regression budget. We evaluate GRASP across five base models on two FHIR-based clinical benchmarks, which score procedural reliability against FHIR state rather than clinical correctness or patient outcomes. On MedAgentBench, GRASP lifts gpt-oss-120b from 40.6% to 88.8%, exceeds the strongest of five self-improvement baselines by 21.0 points, and improves every other base model by 17.2 to 40.3 points. Ablations attribute the gain to comparative proposal generation, the acceptance gate, and the hard regression budget rather than to skill writing itself, which without validation is no better than using no skills. Granting the same acceptance gate to all five baselines lifts each of them in-domain and none of them out of distribution, isolating the gain to the gate applied to a bounded, editable library rather than to held-out validation itself. The mechanism helps in non-clinical environments where tasks recur with verifiable structure and is flat where the action space is open-ended. Frozen libraries transfer across models and across benchmarks that share a tool-calling convention and degrade under interface mismatch.