Search papers, labs, and topics across Lattice.
Graphectory Viewer is a web-based tool designed for the interactive analysis of software-agent trajectories, transforming raw data into phase-aware graphs that link execution details with behavioral structures. This tool enhances the ability to inspect individual agent executions, identify behavioral patterns, and analyze large trajectory datasets, thereby facilitating a deeper understanding of agent performance beyond mere task outcomes. By making this tool open-source and providing comprehensive documentation and resources, the authors aim to support reproducibility and foster further research in the field.
Graphectory Viewer reveals hidden patterns in agent behavior, enabling researchers to dissect software trajectories like never before.
We present Graphectory Viewer, a web-based tool for interactive, process-centric analysis of software-agent trajectories. Building on the Graphectory representation introduced in our previous work, Graphectory Viewer transforms heterogeneous raw trajectories into phase-aware graphs that connect low-level execution details with higher-level behavioral structures. The tool supports trajectories from multiple agent frameworks and provides interactive graph construction; node-level inspection of thoughts, actions, and observations; search and filtering over large trajectory collections; and Sankey-style summaries of problem-solving phase transitions. These capabilities enable researchers and practitioners to inspect individual executions, identify recurring behavioral patterns, compare successful and failed runs, and analyze large trajectory corpora beyond final task outcomes. To support reproducibility and further research, we release Graphectory Viewer as an open-source artifact together with documentation, precomputed graphs, and the large-scale trajectory corpus.
