Project Overview
Trending2Tweet is an autonomous AI agent and content generation pipeline. It systematically identifies high-velocity open-source repositories on GitHub, analyzes repository READMEs and code structure, and drafts clear, insightful technical tweets and breakdown threads for developers on X (Twitter).
Built to demonstrate end-to-end automation, LLM orchestration, and robust API handling in production.
Core Capabilities & Architecture
1. Trend Detection Engine
- Queries the GitHub REST/GraphQL API to extract daily/weekly trending repositories across designated topics and languages.
- Implements intelligent deduplication logic and historical SQLite/JSON tracking to ensure no repository is featured more than once within a 90-day cooldown window.
2. Multi-Provider LLM Orchestration
- Dynamic fallback architecture supporting OpenAI (GPT-4o), Groq (Llama 3), MiMo, and local Ollama models.
- Enforces strict tweet character constraints (280 chars) or multi-part threads while preserving technical accuracy, key benefits, and code snippets.
3. Automated Media & Obsidian Vault Sync
- Generates rich preview cards and syncs all generated drafts and publishing metrics directly into an Obsidian markdown vault for automated knowledge indexing.
Key Metrics & Engineering Wins
- ⚡ Zero-Touch Execution: Fully automated via cron schedules with automatic retries and exponential backoff for rate-limited third-party APIs.
- 🛡️ Cost Optimization: Leveraged Groq and open-weight models for preliminary filtering, slashing LLM API operating costs by 70%.
- 📈 High Engagement: Consistently produces high-signal, fluff-free technical summaries with strong viral appeal in the open-source community.
Tech Stack
- Language: Python 3.12
- AI & LLM: OpenAI API, Groq SDK, LangChain / Prompt Engineering
- Integrations: GitHub REST API, X API v2, Obsidian Vault
- Tooling: Pytest, Pydantic, Dotenv, Rich CLI