Setting Up Nanobot with Telegram Blog Publishing

How I set up nanobot, my AI assistant, to publish blog posts directly via Telegram for frictionless content creation.

I've been working on nanobot, a personal AI assistant that lives in my terminal, and recently added one of my favorite features: the ability to publish blog posts directly through Telegram.

What is Nanobot?

Nanobot is an AI assistant I built that:

  • Stores long-term memories about me (preferences, project details, personal info)
  • Manages daily notes and tasks
  • Has extensible "skills" that can do complex tasks
  • Integrates with services like GitHub, Notion, Apple Notes, and more

The best part? It lives entirely in my terminal and can be controlled naturally through conversation.

The Telegram Integration

The key feature I wanted was the ability to create and publish blog posts without leaving Telegram. I'm often on my phone when ideas strike, and having to:

  1. Open a laptop
  2. SSH into my server
  3. Write the post
  4. Build and push
  5. Wait for Cloudflare Pages

...was just too much friction.

How It Works

The workflow is now incredibly simple:

  1. Telegram: I send a message like "Write a blog post about X"
  2. Nanobot: Receives it, uses the blog-post skill to create the post with proper frontmatter
  3. Validation: Runs a local build to catch any errors
  4. Git Push: Commits and pushes to the main branch
  5. Deploy: Cloudflare Pages automatically builds and deploys

The entire process takes about 30 seconds from idea to live post.

Technical Details

The blog-post skill is a custom nanobot skill that:

  • Uses git to sync with the blog repository
  • Generates proper YAML frontmatter (title, date, category, description, slug)
  • Validates posts with a local python3 build.py before committing
  • Pushes to Cloudflare Pages which auto-deploys

Here's the skill configuration:

name: blog-post
description: Full-stack blog publisher to create content with frontmatter, validate builds, and push to git for Cloudflare Pages deployment.

Frontmatter Example

Each post gets this YAML header automatically:

---
title: "Your Title Here"
date: YYYY-MM-DD
category: [Projects, Learning, Meta]
description: Max 160 chars summary for SEO.
slug: url-friendly-slug

Valid categories include: DevOps, ML-AI, Projects, Career, Learning, Side-Projects, Meta

Why This Matters

This setup has increased my blogging frequency significantly because:

  1. Zero friction: No need to open an editor or remember build commands
  2. Immediate capture: Ideas get captured immediately when they occur
  3. Reliable: The build validation catches errors before they reach production
  4. Fast: Cloudflare Pages deploys in 1-2 minutes

What's Next

I'm planning to add:

  • Image uploads via Telegram for blog post images
  • Draft mode (save without publishing)
  • Analytics integration to see which posts perform best

Wrapping Up

If you're interested in building your own AI assistant with custom skills, check out the nanobot project. The skill system makes it easy to add functionality like blog publishing, GitHub management, or any other automated workflow you can imagine.

Want to see nanobot in action? Just wait for the next blog postโ€”I'll probably write it from Telegram. ๐Ÿ˜„