Turn last week’s GitHub activity into a blog post

An automated pipeline turns recent code activity into a first-person weekly draft and stages it for review in Notion and DEV.

Done with Custom agent· Intermediate· Mar 2026

What they did
The developer created DevNotion with Mastra and scheduled it to run each Sunday through GitHub Actions. One stage pulls seven days of repository activity from GitHub’s GraphQL API, including commits, pull requests, issues, reviews, and language stats, and it does that without an LLM. Gemini then shapes those details into a first-person article, with a fallback path if that step fails, and a publishing stage creates a DEV.to draft alongside a Notion planning page with tables.
How it went
The workflow reliably produces one draft every week, which the author edits before publishing. No audience metrics were reported. Of the models tried, Gemini delivered the strongest output, beating Ollama and Groq-based options.
Tips
Gemini’s native structured output slowed each call by roughly 20–40 seconds. Parsing plain text with YAML frontmatter was 3–4 times faster, and Notion needed rate limiting.

Try it yourself with n8n

Create a weekly automation for me. Look at the past 7 days of activity for [GitHub repo or username], then write a concise summary of what changed and turn it into a first-person post in [tone]. Prepare versions for [Notion] and [DEV], and store a draft in Notion. Stop and ask for my approval before any public publishing or submission happens.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on dev.to ↗

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.