Harmful skills in open agent ecosystems raise average harm scores from 0.27 to 0.76 across six LLMs by lowering refusal rates when tasks are presented via pre-installed skills.
hub Canonical reference
arXiv preprint arXiv:2603.02176 , year=
Canonical reference. 88% of citing Pith papers cite this work as background.
18
Pith papers citing it
Background
88% of classified citations
hub tools
citation-role summary
background 7
other 1
citation-polarity summary
years
2026 18representative citing papers
SkillGenBench is a benchmark for evaluating LLM skill generation pipelines in task-conditioned and task-agnostic regimes from repository and document sources using execution-based checks.
OLIVIA treats LLM agent action selection as a contextual linear bandit over frozen hidden states and applies UCB exploration to adapt online, yielding consistent gains over static ReAct and prompt-based baselines on four benchmarks.
SearchSkill improves exact match scores and retrieval efficiency on open-domain QA by conditioning LLM actions on skills from an evolving SkillBank updated from failure patterns via two-stage SFT.
CMIB uses a conditional multimodal information bottleneck to create reusable agent skills that separate verbalizable text content from predictive perceptual residuals, improving execution stability.
SkillRet benchmark shows fine-tuned retrievers improve NDCG@10 by 13+ points over prior models on large-scale skill retrieval for LLM agents.
A systematic study across five domains finds model-generated skills yield average gains but non-uniform negative transfer, with a meta-skill improving extraction quality.
A meta-skill authors and refines prose-and-code skills for agents by learning from post-deployment failures with an overfit audit, achieving 56.8% accuracy on SkillsBench tasks versus 43.6% for human-curated skills.
SkillRAE organizes skills into a graph and compiles compact, grounded contexts for LLM agents, yielding 11.7% gains on SkillsBench over prior RAE methods.
Skill-R1 applies bi-level group-relative policy optimization to evolve skills recurrently from verified outcomes, yielding gains over baselines on multi-step tasks.
GoSkills converts flat skill lists into role-labeled execution contexts via anchor-centered groups and graph expansion, preserving coverage and improving rewards on SkillsBench and ALFWorld under small skill budgets.
SkillSynth uses a scenario-mediated skill graph to sample workflow paths and generate executable terminal tasks, enabling controlled diversity in training trajectories for agents.
SkillsVote is a governance system for agent skills that profiles corpora, recommends via search, and gates updates on successful reusable outcomes, yielding benchmark gains without model changes.
Skill1 trains a single RL policy to co-evolve skill selection, utilization, and distillation in language model agents from one task-outcome reward, using low-frequency trends to credit selection and high-frequency variation to credit distillation, outperforming baselines on ALFWorld and WebShop.
Bilevel optimization with outer-loop MCTS for skill structure and inner-loop LLM refinement improves agent accuracy on an operations-research question-answering dataset.
LLM agent progress depends on externalizing cognitive functions into memory, skills, protocols, and harness engineering that coordinates them reliably.
citing papers explorer
-
SkillGenBench: Benchmarking Skill Generation Pipelines for LLM Agents
SkillGenBench is a benchmark for evaluating LLM skill generation pipelines in task-conditioned and task-agnostic regimes from repository and document sources using execution-based checks.
-
OLIVIA: Online Learning via Inference-time Action Adaptation for Decision Making in LLM ReAct Agents
OLIVIA treats LLM agent action selection as a contextual linear bandit over frozen hidden states and applies UCB exploration to adapt online, yielding consistent gains over static ReAct and prompt-based baselines on four benchmarks.
-
SearchSkill: Teaching LLMs to Use Search Tools with Evolving Skill Banks
SearchSkill improves exact match scores and retrieval efficiency on open-domain QA by conditioning LLM actions on skills from an evolving SkillBank updated from failure patterns via two-stage SFT.
-
SkillRet: A Large-Scale Benchmark for Skill Retrieval in LLM Agents
SkillRet benchmark shows fine-tuned retrievers improve NDCG@10 by 13+ points over prior models on large-scale skill retrieval for LLM agents.
-
From Raw Experience to Skill Consumption: A Systematic Study of Model-Generated Agent Skills
A systematic study across five domains finds model-generated skills yield average gains but non-uniform negative transfer, with a meta-skill improving extraction quality.
-
SkillEvolver: Skill Learning as a Meta-Skill
A meta-skill authors and refines prose-and-code skills for agents by learning from post-deployment failures with an overfit audit, achieving 56.8% accuracy on SkillsBench tasks versus 43.6% for human-curated skills.
-
Toward Scalable Terminal Task Synthesis via Skill Graphs
SkillSynth uses a scenario-mediated skill graph to sample workflow paths and generate executable terminal tasks, enabling controlled diversity in training trajectories for agents.
-
Skill1: Unified Evolution of Skill-Augmented Agents via Reinforcement Learning
Skill1 trains a single RL policy to co-evolve skill selection, utilization, and distillation in language model agents from one task-outcome reward, using low-frequency trends to credit selection and high-frequency variation to credit distillation, outperforming baselines on ALFWorld and WebShop.
-
Bilevel Optimization of Agent Skills via Monte Carlo Tree Search
Bilevel optimization with outer-loop MCTS for skill structure and inner-loop LLM refinement improves agent accuracy on an operations-research question-answering dataset.
- From Context to Skills: Can Language Models Learn from Context Skillfully?