Fliqr AI documentation instructions
This is the Fliqr AI docs site built with Mintlify. Product: Fliqr AI — https://fliqr.ai · App: https://app.fliqr.ai For Mintlify product knowledge (components, configuration, writing standards), use the installed Mintlify skill and MCP servers. Prefer Mintlify docs MCP over guesswork.About this project
- Pages are MDX (
.mdx) with YAML frontmatter - Site config lives in
docs.json - Scraped source KB lives in
sources/raw-kb/(Chatrace-origin content — rewrite, do not publish raw) - Rewrite tracking:
sources/site/kb-index.json+sources/site/rewrite-progress.json sources/is mintignored and must not appear in the published site
Terminology
Always rebrand source material: replace Chatrace product names, domains, signup links, and community links with Fliqr AI equivalents.
Style preferences
- Active voice, second person (“you”)
- Concise sentences — one idea per sentence
- Sentence case for headings
- Bold UI labels: Click Settings → Channels → Instagram
- Code formatting for paths, commands, file names, and API fields
- No marketing filler, emoji, or hype
- Prefer practical steps a user can complete in the product today
Mintlify usage
When creating or editing pages:- Read
docs.jsonbefore changing navigation - Use Mintlify components where they improve clarity:
<Steps>,<Step>,<Note>,<Tip>,<Warning>,<Info>,<Card>,<CardGroup>,<Accordion>,<AccordionGroup>,<Tabs>,<Tab>,<CodeGroup> - Every page needs
titleand usuallydescription+sidebarTitle - Internal links are root-relative without extensions:
/channels/instagram/story-replies - Add new pages to
docs.jsonor they will not appear - Run
mint validate/mint broken-linkswhen changing structure - Do not invent screenshots; omit or leave a clear
[SCREENSHOT: …]placeholder only when necessary
Content boundaries
- Document customer-facing product features in Fliqr AI
- Do not publish scraped Chatrace pages as-is
- Do not document internal admin-only tooling unless explicitly requested
- Skip Whitelabel entirely — Fliqr AI does not offer or document whitelabel; mark those KB articles
skippedand do not publish pages - Defer cross-category topics (e.g. Ads conversion events) to the correct category instead of forcing them into the current channel
KB rewrite workflow
Work category by category. For each source article insources/site/kb-index.json:
- Read the scraped markdown at
rawPath - Decide docs placement under the Fliqr IA (
channels/,core-concepts/, etc.) - Create or update MDX pages in Fliqr voice with Mintlify components
- Wire navigation in
docs.json(use nested groups when a topic needs subpages) - Update tracking fields on the article:
rewriteStatus:pending|done|deferred|skippeddocsPath: published docs path (no leading slash, no.mdx)rewrittenAt: ISO-8601 timestamprewriteNotes: short note if deferred/skipped or partially mapped
- Refresh rollups in
sources/site/rewrite-progress.json
When to add subpages
Add a nested group when a topic has multiple distinct tasks (setup, story replies, comments, menus, etc.). Keep a short hub/overview page plus one page per task. Prefer depth over stuffing multiple unrelated procedures into one page.Quality bar for rewrites
- Teach the Fliqr workflow, not a literal Chatrace UI clone
- Keep Meta/platform constraints accurate (24h window, Professional account, etc.)
- Link to related Fliqr concepts (
/core-concepts/flows,/core-concepts/ai-agents, etc.) - Mark duplicate source articles carefully — one Fliqr page can cover multiple source URLs if they are the same feature
Agent skills in this repo
.cursor/rules/fliqr-docs.mdc— always-on writing/navigation rules.mintlify/skills/fliqr-docs-writer/SKILL.md— rewrite + Mintlify authoring skill- Root
skill.md— product capability summary for external agents