kevanwee

Chronology — Data skill for Claude Code

Data community

Litigation chronologies as data: mandatory pincites, merge across sources preserving provenance, date-conflict and evidential-gap detection.

How to install Chronology

This entry records only its repository, not the path inside it, so there is no exact command to give. Open kevanwee/chronology and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Chronology does

Litigation chronologies as data: mandatory pincites, merge across sources preserving provenance, date-conflict and evidential-gap detection. CLI and Claude skill.

Alternatives in Data

Privacy & Data Handling — This repository contains: - a plugin package (plugins/compound-engineering) made of markdown/config content 10.8k ★
A Stock Data — A股全栈数据工具包 · 十一层架构 · 54端点 · 19数据源 · 零鉴权 Full-stack China A-share data toolkit for AI agents — 11 layers, 54 end 9.4k ★
Global Stock Data — US stock market data for AI coding assistants — zero-auth, official sources 1.5k ★

README

chronology

Litigation chronologies as data. A schema where every entry carries a pincite, merge across sources that preserves who-said-what, automatic detection of date conflicts and evidential gaps, markdown/CSV rendering, and a Claude skill for the extraction step.

- id: site-meeting
  date: 2024-09-03
  event: Site meeting at Tuas; Alpha's engineer inspected the rejected units.
  actors: [Tan Ah Kow, Lim Bee Hoon]
  sources:
    - { doc: Meeting minutes, bundle_ref: AB-70 }
  event_key: tuas-inspection

$ chronology lint examples/from-correspondence.yaml examples/from-affidavit.yaml
CONFLICT: tuas-inspection: 03 Sep 2024 per Meeting minutes, AB-70 (site-meeting); 05 Sep 2024 per Beta Pte Ltd (inspection-per-lim)
0 warning(s), 1 conflict(s)

Why

Chronologies are built by hand, in Word tables, by the most junior person on the matter, from several sources at once. Three things go wrong: entries lose their page references during editing; two accounts of the same event get silently reconciled into one date; and nobody notices the eight-month stretch with no documents. This repo makes all three mechanical.

It is small on purpose. The schema and the three checks are the value.

What is in the box

Piece	What it does
`schema/chronology.schema.json`	Generated JSON Schema. Entries require at least one source; precision rules are enforced (month precision means the 1st, year means 1 Jan).
`chronology merge`	Unions several files. Same id, or same date + same wording, becomes one entry with all sources. Different dates for the same `event_key` are not merged.
`chronology conflicts`	Same `event_key`, different dates: shows each version and who asserts it. Exit 1 if any.
`chronology gaps --days N`	Stretches of N+ days with no day-precision entry.
`chronology lint`	Sources without a pincite or bundle ref; disputed entries without `asserted_by`; approximate dates without an explanation; events phrased