Get Started
Home
Topics
Search
Library
Research questionHow can computer-use agents retain, refine, and reliably reuse procedural skills across repeated GUI tasks?GUI interaction experience is typically transient, so procedures learned during one rollout may not remain available or reliable on later tasks. Repeated skill revisions can also produce retrieval across tasks and revision churn, making long-term behavior difficult to interpret.
AI
AI Agents
AI Memory
Evaluation & Benchmarks
Latest papersRecent research connected to this question, newest first.From Interaction Traces to Persistent Skills: Online Evolution for Computer-Use AgentsThe evidence covers four OSWorld application domains using the same fixed action-generation and GUI-grounding stack, task sets, and iteration horizons. It compares an evolving skill library with an empty-library control after a five-iteration warm-up; the evolving system achieved higher post-warm-up mean evaluator scores in all four domain runs, while GIMP analysis showed cross-task retrieval and revision churn without guaranteed recovery of the originating task.research paper · Sep 4, 2026
Related questions
How can LLM agents reuse procedural skills across diverse long-horizon tasks without generic or instance-bound memory?How can agent skills support reliable procedural execution while making failures easy to diagnose and repair?How can computer-use agents efficiently coordinate GUI and CLI actions over shared application state?How can agent runtimes avoid context poisoning and latency from growing histories during long-horizon skill execution?