Software Engineering · arXiv:2603.15401 v2 · submitted 16 Mar 2026
SWE-Skills-Bench: Do Agent Skills Actually Help in Real-World Software Engineering?
The paper says
“Our results show that skills provide limited benefits: 92 of 98 skills have no statistically significant effect, and adding skills raises the average pass ratio only from 76.6% to 76.9%.”
from the abstract · picked out as the main claim by a classifier
Reviews
Reviews by people
No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.
The paper
Abstract
Agent skills, structured procedural knowledge packages injected at inference time, are increasingly used to augment LLM agents on software engineering tasks. However, their real utility in end-to-end development settings remains unclear, and no standard procedure tests an individual skill before installation. We present swe-skills-bench, a skill-centric, requirement-driven benchmark that measures how much individual agent skills improve performance on real-world software engineering (SWE) tasks. We start from 98 public SWE skills, each paired with a real GitHub repository pinned at a fixed commit. For each skill, we derive repository-specific requirements, create executable verifiers, and build a reproducible environment. Under each of four agent-model configurations, we run every task once without the skill and once with it, yielding 3,920 paired evaluations. For each requirement, we compute a paired delta, i.e., the change in outcome when the skill is added. We then aggregate these deltas across a skill's ten requirements into a per-skill utility profile, and use a statistical test to classify each skill as Golden (significantly helpful), Ineffective (no significant effect), or Harmful (significantly detrimental). Our results show that skills provide limited benefits: 92 of 98 skills have no statistically significant effect, and adding skills raises the average pass ratio only from 76.6% to 76.9%. The few helpful skills supply procedures that the agent would otherwise omit (gains of up to + 17.7 points). In contrast, harmful skills anchor the agent on their own templates and override the task's explicit instructions (losses of up to -17.0 points). Our code is available at https://github.com/ControllableModel/SWE-Skills-Bench.
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. No sentence in this record was written by a model.
abstract compiled · 2026-10-28 · no language-model review
How the main claim is stated: asserted 1.00
What the abstract itself states, as probabilities: claims first 0.29 · states limitation 0.50 · human participants 0.03 · controlled comparison 0.97 · randomized assignment 0.03 · states quantitative result 0.99 · compares with prior results 0.64 · mentions code or data availability 0.98
The abstract, sentence by sentence
This table labels the abstract only (arXiv listing abstract v2). The number is the probability the classifier gave the label.
| Sentence | Role | The authors' own work | States the main conclusion | Strength |
|---|---|---|---|---|
| Agent skills, structured procedural knowledge packages injected at inference time, are increasingly used to augment LLM agents on software engineering tasks. | background 1.00 | 0.04 | 0.02 | asserted 0.98 |
| However, their real utility in end-to-end development settings remains unclear, and no standard procedure tests an individual skill before installation. | gap 1.00 | 0.05 | 0.03 | asserted 0.89 |
| We present swe-skills-bench, a skill-centric, requirement-driven benchmark that measures how much individual agent skills improve performance on real-world software engineering (SWE) tasks. | contribution 1.00 | 0.95 | 0.06 | asserted 1.00 |
| We start from 98 public SWE skills, each paired with a real GitHub repository pinned at a fixed commit. | method 0.95 | 0.89 | 0.02 | asserted 0.99 |
| For each skill, we derive repository-specific requirements, create executable verifiers, and build a reproducible environment. | method 1.00 | 0.95 | 0.02 | asserted 0.98 |
| Under each of four agent-model configurations, we run every task once without the skill and once with it, yielding 3,920 paired evaluations. | method 1.00 | 0.97 | 0.02 | asserted 0.98 |
| For each requirement, we compute a paired delta, i.e., the change in outcome when the skill is added. | method 1.00 | 0.96 | 0.02 | asserted 0.98 |
| We then aggregate these deltas across a skill's ten requirements into a per-skill utility profile, and use a statistical test to classify each skill as Golden (significantly helpful), Ineffective (no significant effect), or Harmful (significantly detrimental). | method 1.00 | 0.96 | 0.05 | asserted 0.95 |
| Our results show that skills provide limited benefits: 92 of 98 skills have no statistically significant effect, and adding skills raises the average pass ratio only from 76.6% to 76.9%. | result 1.00 | 0.97 | 0.88 | asserted 1.00 |
| The few helpful skills supply procedures that the agent would otherwise omit (gains of up to $+17.7$ points). | result 0.81 | 0.95 | 0.11 | asserted 1.00 |
| In contrast, harmful skills anchor the agent on their own templates and override the task's explicit instructions (losses of up to -17.0 points). | result 0.59 | 0.95 | 0.10 | asserted 1.00 |
| Our code is available at https://github.com/ControllableModel/SWE-Skills-Bench. | contribution 0.88 | 0.39 | 0.02 | asserted 0.93 |
What kind of paper the classifier reports this is: dataset or benchmark 0.96 · evidence from observation or experiment 0.99
What was examined, and what was not
Sources read
- arXiv listing abstract v2
Not examined
- full text
- figures and tables
- references and citing sentences
- relations to other papers
Levels run
- P abstract
Not run
- S sections
- T body sentences
- C citations
- R cross-paper relations
How the labels were made
Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-10-28: 1 request, 8,946 input tokens.
Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. This record as JSON
Cited by 7 Pith papers
Reviewed papers on Pith that cite this one. OpenAlex reports about 106 citations worldwide. The full citation record
- From Procedures to Permissions: Learning Precise Subagent Authorization from Agent Skills cs.CL · Nov 2026
- Evolve Where It Matters: Coverage-Guided Self-Evolution of Agent Skills cs.CL · Nov 2026
- When Agent Skills Go Stale: Runtime Behavior and Selective Maintenance in Evolving Environments cs.AI · Nov 2026
- What Does an Agent Need? Bundle-Aware Agent Resource Composition cs.AI · Nov 2026
- Two Birds with One Stone: Skill-Augmented Generation via Unified Retrieval and Utilization cs.CL · Oct 2026
- SkillCT: A Multidimensional Benchmark and Execution-Free Baseline for Agent Skill Assessment cs.AI · Oct 2026
- SkillNet: Create, Evaluate, and Connect AI Skills cs.AI · Feb 2026
How the literature uses it
48 arXiv papers cite this in 81 sentences
Papers citing it · 50 sentences without a confident label
Uses its method or tool 1 paper
-
Following [ref] on evaluating coding-agent performance, we measure and report: Wrap success ($↑$) measures whether the agent's generated integration loads the released checkpoint and completes the benchmark evaluation protocol end-to-end without manual intervention.
Nautilus: From One Prompt to Plug-and-Play Robot Learning · Ablation: from pairwise wiring to reusable extending · 0.88
Uses its data or values 2 papers
-
Controlled drift construction and validation. starts from 49 agent skills in SWE-Skills-Bench [ref] across seven programming domains.
Skill Drift Is Contract Violation: Proactive Maintenance for LLM Agent Skill Libraries · : A Benchmark for Skill Degradation · 0.98
-
SWE-Skills-Bench provides 490 repository-based software-engineering task instances with no-skill and curated-skill settings [ref] .
Agent Skills Can Be Harmful: An Empirical Study of Skill-Induced Failures in LLM Agents · Study Subjects · 0.86
Cites it as support 5 papers
-
SkillMOO [ref] optimizes pass rate and cost jointly, SkillLearnBench [ref] shows that plausible skill text does not imply competence, SWE-Skills-Bench [ref] reports only $+1.2%$ average gain across public SWE skills with many zero-utility or negative cases, and SkillFlow-Bench [ref] finds that stronger settings maintain compact revised skills while weaker settings often fragment into many low-utility files.
Dynamic Agent Skills: A Lifecycle Survey and Taxonomy of Evolving Skill Libraries · Focused libraries often beat comprehensive ones · 0.97
-
The distinction matters because task-relevant information is not monotonically beneficial: removing detail can improve correctness by disrupting misleading lexical cues [ref] , and packaged skill documents may raise token spend without improving performance [ref] .
Can your AI agent be cheaper? Investigating the effects of task specifications on token spend in agentic coding tasks · Related work · 0.99
-
Controlled skill injection produces task-dependent gains and can add substantial token overhead without improving pass rate [ref] .
Terminal Agents: A Survey of AI Agents in Command-Line Environments · Synthesis: Architectural Trade-offs and Attribution Consequences · 0.98
Cites it as background 18 papers
-
Related Work SkillsBench [ref] introduced the first systematic paired evaluation of skills across diverse tasks, and SWE-Skills-Bench [ref] found that many public SWE skills provide little benefit in practical SE tasks.
SkillMOO: Multi-Objective Optimization of Agent Skills for Software Engineering · Introduction · 0.89
-
SkillsBench [ref] measures structured skills across diverse tasks, while SWE-Skills-Bench [ref] studies the marginal utility of skill documents in real-world software-engineering settings.
Demystifying Agent Skills: Why They Work-Until They Don't · Benchmarks for Evaluating Agent Behavior · 0.83
-
Agent skills have recently emerged as a promising mechanism for retaining and transferring experience, by packaging reusable procedural knowledge as structured instructions for LLM-based agents [ref] .
Do Personalized Skills Help Coding Agents? An Empirical Study of Developer Interaction Traces · Introduction · 0.93
Challenges and discussion
Know something about this paper? Leave a comment.
Come back by email
A short list of the week's newly reviewed papers. Nothing is sent until you confirm from your inbox, and every email has a link that stops it. No account.
No comments yet.
Sign in to comment