Utilities

How to teach the assistant your business context

quattr-context-builder a saved expert workflow your assistant follows, in Claude, ChatGPT or Cursor

interview you and write (or update) the tenant context doc, the per-account defaults every other skill honors: default segment and trackers, priority markets and devices, named competitors, internal jargon mapped to Quattr entities, reporting calendar, KPI priorities. Every id validated live before it's written.

Say any of these

• set up our context
• remember our defaults
• create the tenant context doc
• map our terminology
• make this account's preferences stick
• configure defaults for this client

You never name the skill or a tool, the assistant routes on what you asked. These are the phrasings it recognises for this one.

And when it stays out of the way

What it runs

• describe_data_sources
• resolve_entity
• list_dashboards
• list_filter_values

4 governed analyses, composed and scoped for you. Named here for transparency, not because you have to ask for them.

What it refuses to do

• Nothing enters the doc unvalidated; vocabulary mappings show both your term and the bound id, so drift is auditable.
• The doc records DEFAULTS and language, never numbers, claims or analysis conclusions (those go stale; entities don't). Anything that ranks priorities, states constraints, or records interventions belongs in the companion docs, never in quattr-context.md.
• The companion docs are opt-in by activation rule: operating brief only for prioritization/executive/recurring asks or an explicit ranking request; action ledger only for recurring reports or explicit follow-through management. First-run analysis requires neither.
• An expired operating brief is flagged stale, never silently applied. The ledger governs attribution: no recurring report implies a change caused a movement unless the ledger shows that change shipped.
• On update, changed lines are shown before saving; every doc carries its as-of date.

What the answer looks like

Claude ChatGPT Cursor acme.example · connected

Remember our defaults.

Analysis plan, every step, resolved for you

• Routed to the governed workflowquattr-context-builder
• Dates bound explicitlyJul 1 to Jul 31, 2026
• 3 analyses queued in parallelresolve_entity · list_filter_values · describe_data_sources
• Honesty gates armedscope echo · freshness

The tenant context doc is written, the per-account defaults every other skill honors.

Context doc

written default segment, trackers, markets, KPI order

Captured

1 segment · 3 competitors · 2 markets plus your jargon mapped to Quattr entities

IDs validated live

9 of 9 nothing written that didn't resolve

Jul 1 to Jul 31, 2026

⚠ Observationalas of Aug 1, 2026IllustrativeOpen in Quattr ↗

Next session starts scoped, no re-asking.

A simulated conversation on fictional data. Scope, freshness and the ⚠ Observational marker are attached by the same director that renders every demo here, not written by hand.