# Do LLM-Generated Skills Make Better AI Data Scientists? A Component Ablation Across Data-Science Workflows

arXiv: https://arxiv.org/abs/2607.07504 · HTML: https://mlsanity.com/paper/2607.07504/

## Key facts

- Cited by 3 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 12 of its 15 references that are in the map.
- First posted on arXiv 8 Jul 2026.

## Placed in

- LLM agents and tool use / [Agent skill learning and evolution](https://mlsanity.com/agents/agent-skill-learning-and-evolution/)

## Builds on (most cited first, 12 of 12)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?](https://mlsanity.com/paper/2202.12837/) (2022)
- [Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Task](https://mlsanity.com/paper/1809.08887/) (2018)
- [Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivity](https://mlsanity.com/paper/2104.08786/) (2021)
- [DS-1000: A Natural and Reliable Benchmark for Data Science Code Generation](https://mlsanity.com/paper/2211.11501/) (2022)
- [SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks](https://mlsanity.com/paper/2602.12670/) (2026)
- [SkillLearnBench: Benchmarking Continual Learning Methods for Agent Skill Generation on Real-World Tasks](https://mlsanity.com/paper/2604.20087/) (2026)
- [AutoDCWorkflow: LLM-based Data Cleaning Workflow Auto-Generation and Benchmark](https://mlsanity.com/paper/2412.06724/) (2024)
- [SkVM: Revisiting Language VM for Skills across Heterogenous LLMs and Harnesses](https://mlsanity.com/paper/2604.03088/) (2026)
- [LLM4DS: Evaluating Large Language Models for Data Science Code Generation](https://mlsanity.com/paper/2411.11908/) (2024)
- [Chain-of-Query: Unleashing the Power of LLMs in SQL-Aided Table Understanding via Multi-Agent Collaboration](https://mlsanity.com/paper/2508.15809/) (2025)
- [PrepBench: How Far Are We from Natural-Language-Driven Data Preparation?](https://mlsanity.com/paper/2605.08687/) (2026)

## Built on by (most cited first, 3 of 3)

- [Is Progressive Disclosure All You Need for Long-Context Agents?](https://mlsanity.com/paper/2607.17598/) (2026)
- [An Empirical Study of Agent Skills' Downstream Utility](https://mlsanity.com/paper/2610.08875/) (2026)
- [Cortex: Measurement-Gated Evolution and Pruning of Coding-Agent Skills and Rules](https://mlsanity.com/paper/2610.14216/) (2026)

## Abstract (arXiv)

Product data scientists often ask LLM-based agents to help with recurring execution tasks such as cleaning data, writing SQL, choosing statistical tests, and formatting results. Reusable skill files are meant to avoid prompting from scratch by packaging guidance for a task family. Expert-written skills can encode high-quality guidance, but writing and maintaining them across many data-science task families creates a manual bottleneck. We ask whether LLM-generated skills offer a useful low-curation alternative: do they improve performance over the task prompt alone? We test this question across four lifecycle stages: data preparation, data extraction, statistical analysis, and reporting, using one generated skill per stage. We find no reliable improvement from full generated skills over No-Skill prompting. We then ask whether any part of the skill is useful by ablating different skill components. The main ablation covers 56 tasks, nine model configurations, and three providers, yielding 7,560 runs. Compared with prompting using the task alone, neither the full generated skill nor any ablated skill variant significantly improves performance; all p-values are at least 0.396, and the total spread across variants is only 1.2 pp. A supplemental token-matched control adds 1,512 runs and finds that Full skills perform similarly to task-irrelevant skill-formatted content. The results caution against using one LLM-generated skill per data-science workflow as a default single-shot prompting strategy.