AI-powered daily summaries of your X/Twitter feed: scraped, summarized, narrated, and delivered by email.
This repository is the new home of FeedTLDR: a Next.js + shadcn/ui frontend and a FastAPI layer wrapping the existing Python pipeline (Apify scraping, Gemini summarization, TTS audio, SendGrid newsletters, Firebase, Stripe). It replaces the Streamlit app in feedtldr_streamlit.
- docs/PLAN.md — rebuild plan: current-state analysis, architecture, API contract, phased roadmap
- docs/DESIGN.md — design system: tokens, typography, components, page specs
Status: Phases 0–5 implemented (backend re-homed + FastAPI layer + full Next.js frontend, all suites green). Remaining before cutover: real credentials/env wiring, live end-to-end QA (PLAN.md Phase 6), and deployment (Phase 7).
# API (needs apps/api/.env + credentials/, see .env.example)
cd apps/api && uv sync --all-extras && uv run uvicorn api.main:app --reload
# Web (needs apps/web/.env.local, see .env.example)
cd apps/web && pnpm install && pnpm dev