How to write and maintain the documentation in this repository so it stays accurate, consistent, and useful to AI coding agents and humans alike.
This repository is public. Never add secrets, credentials, private hostnames, IP addresses, or tenant identifiers to any file.
../AGENTS.md — top-level agent guide and index (kept under 150 lines).../CLAUDE.md — a pointer file that imports AGENTS.md. Do not add content
here; edit AGENTS.md instead.tech/ — technical setup, tooling, and CI guides, indexed by tech/README.md.domain/ — data model and glossary, indexed by domain/README.md.snowflake-data-inventory.md — data conventions, access options, and the table index.snowflake-tables/ — one reference document per Snowflake table.qliksense-reports/, reporting-general/ — reporting and KPI documentation.CLAUDE.md pointer file.#/##/### headings; do not nest deeper than ### in sub-files.tech/ and domain/ sub-files. (The legacy landing-page and inventory files use
emojis in section headers; leave those as-is unless reworking the page.)AGENTS.md under 150 lines and each docs/** sub-file under 300 lines.lower_snake_case.md matching the table name. Guides in tech/ and
domain/ use UPPERCASE-KEBAB-CASE.md.snowflake-tables/.snowflake-data-inventory.md (and ../readme.md if the table is featured).../toc.json so it appears in the Stoplight navigation.stoplight-id front matter on existing files.When platform data or conventions change, update the affected reference file, the glossary
(domain/GLOSSARY.md), and the data model overview (domain/DATA-MODEL.md). Automated code review
checks documentation PRs against the repository and flags stale content.