Every time I build something, I talk to Claude. That conversation becomes a build log. That build log becomes an article. The whole thing is automated.
Here is how the pipeline works.
Step 1: The Build Session
I build everything through Claude conversations. Setting up a VoIP system? That is a conversation. Deploying an edge function? Conversation. Debugging a PJSIP trunk? Long conversation.
Each conversation is a build log. It has the problem, the solution, the code, and the context.
Step 2: Auto-Draft
A Supabase edge function called auto-draft-content takes a session summary and sends it to Claude Haiku with a system prompt:
"You are a technical writer for dev.to. Convert this build session into a practical, first-person technical article. Use short sentences. No fluff. No em dashes. Include code snippets where relevant. Target 800 to 1200 words."
Claude Haiku is fast and cheap. At $0.25 per million input tokens, drafting an article costs less than a penny. The output goes into the drafted_content column of the build_content_queue table.
Step 3: Publish
Another edge function called publish-build-content takes articles with status ready, parses the frontmatter (title, tags), checks for duplicates against my dev.to profile, and publishes via the dev.to API. Rate limiting is built in with 35-second delays between posts.
Step 4: Schedule
A pg_cron job fires at 9 AM Central every day and publishes up to 3 articles. Consistent cadence without manual intervention.
The Results
32 articles published in under 2 weeks. All from real build sessions. Not AI slop. Real problems, real solutions, real code. The AI just handles the formatting and publishing.
The Cost
Claude Haiku for drafting: roughly $0.005 per article
dev.to API: free
Supabase edge functions: free tier
pg_cron: included with Supabase
Total monthly cost for an automated content pipeline: less than $1.
Why This Matters for Your Brand
Every developer has build sessions. Most of them disappear into Slack threads and terminal history. If you capture them and publish them, you build a public record of your work. That record becomes your resume, your portfolio, and your lead generation engine.
I got asked about my content during a job interview. That pipeline is not just content. It is career infrastructure.
Top comments (0)