Instruction file imported from Orenda-Project/rumi-platform (.cursor/rules/rumi.mdc). Copyright stays with the author.

Rumi Platform

Name: rumi
Author: Orenda-Project

Open-source AI teaching assistant on WhatsApp (Node.js + Express + Supabase + BullMQ).

Architecture

bot/whatsapp-bot.js — Entry point (webhook, message routing)
bot/shared/config/ — Branding, feature tiers, capabilities
bot/shared/services/ — 39+ service modules (LLM, coaching, reading, video)
bot/shared/handlers/ — Message handlers (text, voice, image, flow)
bot/workers/ — 8 background workers
infrastructure/supabase/ — SQL schema (52+ tables)
tests/ — 158 tests across 11 suites

Rules

Never commit .env files — use .env.template for documentation
Branding via bot/shared/config/branding.js — no hardcoded bot names or org names
All LLM calls go through bot/shared/services/llm-client.js
Background jobs use BullMQ (Redis), not SQS
Feature gating via RUMI_TIER env var: minimal, recommended, or full
Tests run without external services — no Supabase/Redis/WhatsApp needed

Commands

npm test              # All tests
npm run test:security # Security scan
npm run validate:env  # Environment check
npm run simulate      # CLI simulator

Key Files for Common Changes

Change	Files
Bot personality/name	`bot/shared/config/branding.js`
LLM model/provider	`bot/shared/services/llm-client.js`, `.env`
Feature availability	`bot/shared/config/feature-tiers.js`
Coaching framework	`bot/shared/services/coaching-service.js`, `bot/shared/constants/scoring-rubric.js`
Reading benchmarks	`bot/shared/services/reading-assessment-service.js`
Add a language	7 config files (see `docs/agent-customization.md#4`)

{ "ocm": "1", "id": "orenda-project-rumi-platform-rumi-rule", "kind": "skill", "name": "rumi", "description": "Rumi Platform project rules for WhatsApp AI teaching assistant", "publisher": "Orenda-Project", "version": "1.0.0", "capabilities": { "domains": [ "general" ], "tags": [ "skill-md", "github-cursor-rules" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "Rumi Platform project rules for WhatsApp AI teaching assistant" ], "primary": false, "metadata": { "source": { "provider": "github-cursor-rules", "repository": "https://github.com/Orenda-Project/rumi-platform", "path": ".cursor/rules/rumi.mdc", "ref": "ef1dd7686871451c594a6c1974f98489e67b5a9b", "url": "https://github.com/Orenda-Project/rumi-platform/blob/ef1dd7686871451c594a6c1974f98489e67b5a9b/.cursor/rules/rumi.mdc", "key": "Orenda-Project/rumi-platform/.cursor/rules/rumi.mdc" }, "applies_to": [ "bot/**/*.js", "tests/**/*.js", "infrastructure/**/*.sql" ] }, "instructions": "# Rumi Platform\n\nOpen-source AI teaching assistant on WhatsApp (Node.js + Express + Supabase + BullMQ).\n\n## Architecture\n\n- `bot/whatsapp-bot.js` — Entry point (webhook, message routing)\n- `bot/shared/config/` — Branding, feature tiers, capabilities\n- `bot/shared/services/` — 39+ service modules (LLM, coaching, reading, video)\n- `bot/shared/handlers/` — Message handlers (text, voice, image, flow)\n- `bot/workers/` — 8 background workers\n- `infrastructure/supabase/` — SQL schema (52+ tables)\n- `tests/` — 158 tests across 11 suites\n\n## Rules\n\n1. Never commit `.env` files — use `.env.template` for", "cost": { "context_tokens": 416 } }

rumi

About

Rumi Platform

Architecture

Rules

Commands

Key Files for Common Changes

Use it

Manifest

Reviews