Autopilot - Ctrl: AI Content Auditing with GitHub Copilot CLI
When AI generates social media content, how do we know if it’s good? I built autopilot-ctrl, a tool that uses GitHub Copilot CLI to evaluate content quality before publishing.
When AI generates social media content, how do we know if it’s good? I built autopilot-ctrl, a tool that uses GitHub Copilot CLI to evaluate content quality before publishing.
We add a new feature to the Autopilot: a newsletter integrated into the blog footer that captures subscribers, segments them by language (ES/EN), and automatically sends AI-generated personalized emails when I publish new content.
In this last chapter, we ditch manual execution. We built a CI/CD pipeline in GitHub Actions that detects new articles, orchestrates AI agents, and manages publishing to Twitter and LinkedIn under human supervision. Welcome to total automation.
Connecting an API is usually easy… until you try to post to a LinkedIn Company Page. In this post, I recount the odyssey of permissions, verifications, and ‘Marketing Developer Platform’ forms I had to overcome so my Python script could officially speak on behalf of Datalaria.
Having data isn’t enough; no one likes a JSON file. In this post, we design the personality of our Writer Agents, teach Gemini to write ‘Broetry’ for LinkedIn and ‘Shitposting’ for Twitter, and scale the architecture to publish in Spanish and English simultaneously.
A script that reads files is easy. A script that ‘understands’ technology is a different story. In this post, we configure the Python environment, solve CrewAI integration errors, and succeed in having Gemini Flash extract ‘pure gold’ from our Markdown posts.
We kick off our weather project by building the engine: a Python script that talks to an API, saves historical data, and runs daily thanks to GitHub Actions. I’ll share the tricks and challenges!