Skip to content

Explore a trajectory Dataset

pChronicle reads trajectory Datasets from local directories or S3, discovers supported source formats, and exposes normalized runs, steps, and tool_calls tables.

It is the data layer for captured runtime history and external trajectory files—not the runtime or scheduler. External files are pinned and normalized directly rather than converted into canonical runtime events.

Start with a known Dataset

From a Persisting source checkout:

pchronicle ls examples/data/atif
pchronicle analysis overview examples/data/atif

ls shows the discovered Sources. The overview reports the shape of the Dataset without requiring you to write SQL first.

Ask a specific question

pchronicle query examples/data/atif \
  'SELECT source, COUNT(*) AS steps
   FROM dataset.steps
   GROUP BY source
   ORDER BY source'

Queries are read-only. Format-specific data is normalized at the Dataset boundary so the same question can span ATIF, ACTF, OpenAI Messages, canonical events, and Storyline Sources where their semantics align.

What you completed

You discovered the logical Sources in one Dataset, created a Catalog Snapshot, ran a stable built-in summary, and queried a normalized relation. No Dataset content was modified.

Continue by task: