pChronicle

pChronicle is Persisting's structured trajectory and Dataset data layer. It discovers native and supported external Sources on local storage or S3, preserves canonical event facts where present and keeps Source origin visible, exposes normalized Run views, and supports bounded query, analysis, revision lineage, and format exchange.

What pChronicle owns

• Dataset and Source discovery;
• immutable Catalog Snapshot membership and source version descriptions;
• canonical event storage and terminal Run facts;
• normalized runs, steps, and tool_calls query views;
• import boundaries for ATIF, ACTF, and OpenAI Messages;
• export boundaries for those formats and Storyline JSON;
• AgenticMD as a non-authoritative human-readable projection;
• revision lineage for derived data.

pChronicle does not execute or schedule Agents. Its inputs include canonical runtime-event Sources and pinned local or S3 ATIF, ACTF, OpenAI Messages, and Storyline Sources. External Sources are normalized directly; they do not first become canonical runtime events.

Ask the first question

pchronicle onboard
pchronicle onboard query

These installed-product walkthroughs create temporary example Datasets and do not require a source checkout.

Read pChronicle by purpose

Goal	Section
Query the first Dataset	Get Started
Understand Dataset, Source, events, and projections	Concepts
Follow common trajectory data workflows	Guides
Inspect storage and catalog mechanisms	Design
Look up commands, schemas, and formats	Reference

For Persisting-governed capture through pVisor's configured Gateway and lifecycle-event path, begin with pVisor.