repo-promoter-agent

Inspiration

I maintain several open source projects on GitHub and noticed that promoting them is often harder than building them. I wanted a simple way to turn any repo into ready‑to‑use marketing content so more developers can actually find the projects I work on.

What it does

You paste a public GitHub repo URL into the app, and it generates a complete “promo pack”: a headline, short summary, key benefits, tweet‑sized posts, a LinkedIn‑style post, and a call to action. All generated content is stored and made searchable so you can quickly reuse good copy across multiple repositories.

How I built it

The app uses a small web frontend, a Go backend, a Gradient AI agent, and Apache Solr 10. The backend fetches repo data from GitHub, sends structured input to a Gradient AI agent, gets back JSON‑formatted promotional content, and indexes it in Solr. The frontend has two main screens: “Generate” for creating promo packs and “Search” for browsing everything generated so far.

What I learned

Working with DigitalOcean’s Gradient AI platform was easier than I expected once the agent schema was clear. I learned how important it is to design prompts like contracts so the backend can rely on the JSON output and Solr can index it without surprises.

Challenges

Configuring Solr 10 correctly for this kind of text search took time, especially around the schema and query settings. It was also challenging to pick a model and settings for the Gradient AI agent that consistently produced valid JSON while still sounding natural and marketing‑friendly.

Costs

First day live costed $4.53 - App (29 hours) $1.07, Droplet (34 hours) $1.62, Database (58 hours) $1.69, Agents $0.15. That will increase as I am going to try different Agents and Models.

Built With

• github
• go
• solr