Demystifying Agent Skills: Why They Work-Until They Don't
Paper • 2608.14036 • Published • 163
Papers
-
-
Harness the Memory: A Holistic Evaluation of Memory Substrates in Memory Agents
Paper • 2608.15008 • Published • 15 -
Oracle Agent Memory as an Enterprise Memory Substrate for Long-Horizon AI Agents
Paper • 2607.13157 • Published -
SelfMem: Self-Optimizing Memory for AI Agents
Paper • 2607.03726 • Published -
What to Keep, What to Forget: A Rate--Distortion View of Memory Compaction in LLMs and Agents
Paper • 2607.08032 • Published -
Memory as a Controlled Process: Learned Adaptive Memory Management for LLM Agents
Paper • 2607.13591 • Published -
SkillForge: Self-Distilling Agents for Project-Specific Issue Resolution
Paper • 2608.18933 • Published • 11 -
Know Before Fix: QA-Driven Repository Knowledge Acquisition for Software Issue Resolution
Paper • 2607.11111 • Published • 25 -
FastContext: Training Efficient Repository Explorer for Coding Agents
Paper • 2606.14066 • Published • 96 -
StartupBench: Benchmarking General-Purpose Agents on Market-Validated End-to-End Workflows
Paper • 2608.17800 • Published • 9 -
TUA-Bench: A Benchmark for General-Purpose Terminal-Use Agents
Paper • 2606.28480 • Published • 48 -
Managing Procedural Memory in LLM Agents: Control, Adaptation, and Evaluation
Paper • 2606.23127 • Published • 26 -
SkillAdaptor: Self-Adapting Skills for LLM Agents from Trajectories
Paper • 2606.01311 • Published • 37 -
SkillX: Automatically Constructing Skill Knowledge Bases for Agents
Paper • 2604.04804 • Published • 35 -
GUI vs. CLI: Execution Bottlenecks in Screen-Only and Skill-Mediated Computer-Use Agents
Paper • 2606.24551 • Published • 28 -
TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks
Paper • 2605.22535 • Published • 11 -
Agent Skills Can Be Harmful: An Empirical Study of Skill-Induced Failures in LLM Agents
Paper • 2608.11888 • Published • 1 -
Evo-Harness: Context-to-Harness Skill Compilation for Self-Evolving Agents
Paper • 2608.15071 • Published -
Linear Attention Architectures: Mechanisms, Trade-offs, and Cross-Layer Routing
Paper • 2607.07953 • Published • 16 -
Efficient Reinforcement Learning for Long-Horizon Tool-Use Agentic Tasks
Paper • 2608.10357 • Published -
Harness-R1: Learning to Edit Executable Runtime Harnesses from Agent Failure Trajectories
Paper • 2608.02276 • Published • 4