Examples of work in Core. Each one unfolds over time and keeps its sources, assumptions, alternatives, and decisions in one continuing workstream.
Why a workstream, not a chat
Work continues across sessions instead of resetting.
Sources stay attached and traceable.
Assumptions and alternatives stay visible.
Python where useful, review before it counts, full history.
Individual
Across months
Maintain sources, assumptions, alternatives, calculations, open questions, and decisions across a relocation, major purchase, career change, or education decision.
Across a long project
Preserve research, sources, characters, themes, editorial decisions, unresolved questions, and drafts in one workstream.
With real numbers
Track bids, documents, budget, decisions, change orders, open questions, and family input. Use Python where quantitative analysis helps.
Families and trusted groups
In their own words
Combine interviews, photographs, records, dates, and memories, including conflicting recollections, without flattening every contribution into one fact.
Across the family
Manage responsibilities, schedules, documents, questions, and decisions while preserving individual and shared boundaries.
Together
Travel, estate organization, renovation, or a community project, with a clear record of who decided what.
Private individual understanding does not automatically become group understanding. Group understanding does not automatically become public.
Small teams
One workstream
Bring together founders, finance, model analysis, Python forecasts, source data, assumptions, and reviewer input.
With provenance
Preserve client context, sources, assumptions, analysis, deliverables, handoffs, and review history.
With decision history
Maintain product decisions, research, code work, user feedback, open questions, technical analysis, and decision history.
A workstream brings together people, models, Python workers, APIs, Skills, and reviewers, and keeps track of who did what. Worker output is draft and pending review.