add warp.md for use in the warp terminal

warp.md

@@ -0,0 +1,148 @@
+# WARP.md
+
+This file provides guidance to WARP (warp.dev) when working with code in this repository.
+
+## Repository Overview
+
+This is the Tiger Data documentation repository, containing the source content for https://www.tigerdata.com/docs/. Tiger Data (formerly Timescale) is a modern Postgres data platform for time series, events, analytics, and vector search. The documentation is written in Markdown with custom components and is built separately in a private Gatsby-based repository.
+
+## Development Workflow
+
+### Link Checking
+Run link checker locally:
+```bash
+npx markdown-link-check <file.md> --config mlc_config.json
+```
+
+### Linting Documentation
+The repository uses Vale for prose linting:
+```bash
+vale --glob='["_partials/**/*", "_troubleshooting/**/*", "about/**/*", "api/**/*", "getting-started/**/*", "mst/**/*", "navigation/**/*", "self-hosted/**/*", "tutorials/**/*", "use-timescale/**/*", "ai/**/*"]'
+```
+
+Vale configuration is in `.vale.ini` and uses Vale + Google style guides.
+
+### Markdown Linting
+Custom markdownlint rules are in `.vscode/markdownlint/` and configured via `.markdownlint.json` and `.markdownlint-cli2.jsonc`. These enforce documentation-specific patterns like:
+- Proper blank lines around `<Highlight>` blocks
+- Correct `<Procedure>` formatting
+- Proper frontmatter structure
+
+### Creating Hyperfunction Documentation
+Use the template generator script for new two-step aggregate hyperfunction groups:
+```bash
+npm run template:hyperfunction
+```
+
+This interactive script creates the directory structure and template files in `api/_hyperfunctions/`.
+
+### Bulk Editing API Excerpts
+Helper scripts in `.helper-scripts/` for bulk frontmatter editing:
+```bash
+cd .helper-scripts
+./extract-excerpts.sh  # Extract excerpts to single file
+# Edit extracted_excerpts.md
+./insert-excerpts.sh   # Update original files
+```
+
+## Architecture
+
+### Navigation Hierarchy
+Navigation structure is governed by `page-index/page-index.js` files throughout the repository. The top-level `page-index/page-index.js` imports and combines all section-level navigation files.
+
+**Page index structure:**
+- `href`: URL segment, must match markdown filename (without extension)
+- `title`: Page name in left TOC
+- `excerpt`: Short description (up to 100 chars) for page cards
+- `type`: Optional - `directory`, `placeholder`, `redirect-to-child-page`
+- `children`: Array of child pages
+- `pageComponents`: Display style for child pages - `['featured-cards']` or `['content-list']`
+- `index`: Specify alternative file if not `index.md`
+
+Major documentation sections:
+- `getting-started/` - Getting started guides
+- `use-timescale/` - Product usage documentation
+- `api/` - API references and hyperfunctions
+- `tutorials/` - Step-by-step tutorials
+- `integrations/` - Integration guides
+- `self-hosted/` - Self-hosted deployment docs
+- `mst/` - Timescale MST documentation
+- `ai/` - AI and vector search features
+- `migrate/` - Migration guides
+- `about/` - Company and product information
+
+### Content Organization
+
+**Partials (`_partials/`):** Reusable content snippets. Create files starting with underscore. Import into target `.mdx` pages.
+
+**Troubleshooting (`_troubleshooting/`):** Individual troubleshooting entries (not full pages). Each file contains frontmatter with:
+- `title`: Entry title
+- `section`: Must be `troubleshooting`
+- `products` and/or `topics`: Where entry appears (at least one required)
+- `errors`: Optional error display with `language` and `message`
+- `keywords`: Displayed at bottom, link to related pages
+- `tags`: Affect related page calculations (not displayed)
+
+Troubleshooting pages are programmatically assembled during build. If page doesn't exist, add entry in page-index with `type: "placeholder"`.
+
+**API Documentation (`api/`):** Function references include:
+- Function name with empty parentheses if takes arguments
+- Brief description with warnings
+- Usage samples demonstrating argument syntax
+- Argument table: `Name`, `Type`, `Default`, `Required`, `Description`
+- Return table: `Column`, `Type`, `Description`
+
+Hyperfunctions are organized in `api/_hyperfunctions/` with subdirectories for two-step aggregation pattern functions.
+
+### Custom Components
+
+Available formatting beyond standard Markdown:
+- `<Procedure>`: Step-by-step instructions with bold step summaries
+- `<Highlight>`: Note/warning/caution blocks (use sparingly)
+- Tabs: Multi-option content display
+- Multi-tab code blocks
+- Tags
+
+See `_partials/_formatting_examples.md` for syntax and examples.
+
+### Variables
+Documentation uses variables for product names and UI elements with syntax `$VARIABLE_NAME`. Variables don't work in:
+- Page frontmatter
+- HTML tables and tabs
+
+Internal links: use `:currentVersion:` instead of `latest` in URLs, no need for full domain `https://www.tigerdata.com/docs`.
+
+## Style Guidelines
+
+Follow [Google Developer Documentation Style Guide](https://developers.google.com/style) with exceptions:
+- Do NOT capitalize first word after a colon
+- Use code font (backticks) for UI elements instead of semi-bold
+
+### SEO Optimization
+Include at top of pages:
+- `title`: Up to 60 characters, variation of page title
+- `excerpt`: Under 200 characters, variation of intro
+
+Summarize paragraph content in first sentence. Include main keywords in meta tags, page title, first header, and intro.
+
+## Deployment
+
+### Preview Builds
+PRs from this repository automatically trigger preview builds (allow 10 minutes). Preview URLs follow pattern: `https://docs-dev.timescale.com/docs-<branch-name-with-hyphens>`
+
+Build occurs in private `timescale/web-documentation` repository using Gatsby.
+
+### Production
+Production deployment triggered by changes to appropriate branch.
+
+## Git Workflow
+
+Branch from `latest` for all changes. PRs target `latest` branch.
+
+## Important Notes
+
+- Only commit when explicitly asked by user
+- Never include secrets or sensitive information in documentation
+- Sign CLA on first PR (automated)
+- Large codebases can be excluded from indexing using `.warpindexingignore`
+