# RobustSGPO: Search-Space Control for Agent Harness Evolution

arXiv: https://arxiv.org/abs/2609.09646 · HTML: https://mlsanity.com/paper/2609.09646/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 7 of its 14 references that are in the map.
- First posted on arXiv 9 Sep 2026.

## Builds on (most cited first, 7 of 7)

- [Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution](https://mlsanity.com/paper/2309.16797/) (2023)
- [Illuminating search spaces by mapping elites](https://mlsanity.com/paper/1504.04909/) (2015)
- [Self-Harness: Harnesses That Improve Themselves](https://mlsanity.com/paper/2606.09498/) (2026)
- [Adaptive Auto-Harness: Sustained Self-Improvement for Agentic System Deployment on Open-Ended Task Streams](https://mlsanity.com/paper/2606.01770/) (2026)
- [Harness-Aware Self-Evolving: Co-Evolving Model Weights, Harness, and Task Solutions](https://mlsanity.com/paper/2607.03935/) (2026)
- [AgentX: Towards Agent-Driven Self-Iteration of Industrial Recommender Systems](https://mlsanity.com/paper/2606.26859/) (2026)
- [HELIX: Model-Harness Co-evolution for Recursive Self-Improvement](https://mlsanity.com/paper/2608.13951/) (2026)

## Built on by (most cited first, 1 of 1)

- [Advancing Model Research in AgentX: Long-Horizon Autonomy for Industrial Recommender Systems](https://mlsanity.com/paper/2609.30001/) (2026)

## Abstract (arXiv)

Semantic-gradient-based prompt optimization (SGPO) improves agent harnesses using execution feedback, but its local update rule leaves the choice of edit scope and operation unresolved. We introduce RobustSGPO, which specifies the requested edit, constructs and checks the patch, and continues search from either the incumbent or retained snapshots. We evaluate permission scheduling, cumulative controls, and task-family transfer in the AgentX brainstorming workflow using 120 tasks, 95 runs, and 7,350 candidate attempts. Periodic 1→2→3 scheduling exceeds fixed maximum permission by 0.28 test-score points. RobustSGPO increases completion on 30 held-out tasks from 60.0% to 80.0% and improves test quality from 3.77 to 4.14 under a 20-million-token budget. Category retention reduces source-task degradation after a shift, whereas random retention reaches a higher destination endpoint. Search-space control benefits quality through executable edits and alternative starting points, with measurable retention overhead.