SkillFlow: Scalable and Efficient Agent Skill Retrieval System
cs.AI, cs.CL, cs.MA
Submitted: 2025-04-08
Updated: 2026-09-25
Comments: Accepted to COLM 2026
Code: https://github.com/IBPA/skill-flow
License: http://creativecommons.org/licenses/by/4.0/
The gist: AI agents can extend their capabilities at inference time by loading reusable skills into context, yet equipping an agent with too many skills, particularly irrelevant ones, degrades performance.
Terminology
Abstract
AI agents can extend their capabilities at inference time by loading reusable skills into context, yet equipping an agent with too many skills, particularly irrelevant ones, degrades performance. As community-driven skill repositories grow, agents need a way to selectively retrieve only the most relevant skills from a large library. We present SkillFlow, the first open, multi-stage retrieval system for agent skill discovery that frames skill acquisition as an information retrieval problem over a corpus of 35K community-contributed SKILL.md definitions indexed from GitHub. The pipeline progressively narrows a large candidate set through four stages (dense retrieval, two rounds of cross-encoder reranking, and LLM-based selection), balancing recall and precision at each stage. We evaluate SkillFlow on two coding benchmarks: SkillsBench, a benchmark of 87 tasks and 229 matched skills; and Terminal-Bench, a benchmark that provides only 89 tasks, and no matched skills. On SkillsBench, SkillFlow-retrieved skills raise Pass@1 from 9.2% to 16.4% (+78.3%, p adj = 3.64 times 10-2), reaching 84.1% of the oracle ceiling, while on Terminal-Bench, agents readily use the retrieved skills (70.1% use rate) yet show no performance gain, revealing that retrieval alone is insufficient when the corpus lacks high-quality, executable skills for the target domain. SkillFlow demonstrates that framing skill acquisition as an information retrieval task is an effective strategy, and that the practical impact of skill-augmented agents hinges on corpus coverage and skill quality, particularly the density of runnable code and bundled artifacts. (GitHub: https://github.com/IBPA/skill-flow)
Sources
- Evaluating Large Language Models Trained on Code
- Agent AI: Surveying the Horizons of Multimodal Interaction
- GPT-4o System Card
- SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks
- Passage Re-ranking with BERT
- Multi-Stage Document Ranking with BERT
- OpenAI GPT-5 System Card
- Voyager: An Open-Ended Embodied Agent with Large Language Models
- Text Embeddings by Weakly-Supervised Contrastive Pre-training
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems