Researchers launch Graphectory Viewer for AI agent analysis
New open-source tool turns messy AI agent logs into interactive graphs for debugging complex workflows.
Graphectory Viewer is a web-based tool for interactive, process-centric analysis of software-agent trajectories. It builds on the Graphectory representation to transform 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 offers 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. This enables 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. The tool is released as an open-source artifact with documentation, precomputed graphs, and a large-scale trajectory corpus.
- Graphectory Viewer converts raw AI agent trajectories into interactive, phase-aware graphs for process-centric analysis
- Features include node-level inspection, search/filtering, and Sankey-style phase transition summaries across multiple agent frameworks
- Open-sourced with documentation, precomputed graphs, and a large trajectory corpus for reproducibility
Why It Matters
Finally gives engineers a systematic way to debug and optimize complex AI agent workflows beyond simple outcome metrics.