From Clawdbot to Moltbot: What if we built smart automation from our own tools?

Recently, I came across the GitHub account of a developer who has been steadily building a collection of automation tools since 2009. His projects include everything from controlling macOS interfaces and automating GUIs to running WhatsApp and iMessage clients directly from the terminal. Instead of relying on flashy, finished products, this developer stitched together many small utilities — some unfinished, some weekend projects — into a working ecosystem. This approach caught my attention because it mirrors how real-world innovation often happens: through persistent tinkering rather than overnight success.

For Norwegian SMBs, especially companies with 10 to 50 employees, the automation landscape can feel overwhelming and expensive. Administrative tasks involving systems like Tripletex, Fiken, Vipps, and NAV APIs often require either manual work or costly consultants billing upwards of 400–600 NOK per hour. Specialists can easily demand 1000 NOK or more per hour. This makes the idea of building your own tailored automation seem intimidating and unreachable for many small businesses.

But what if you could piece together your own toolkit, similar to this developer’s approach? Imagine managing your accounting, messaging, and workflow automation with lightweight, modular tools that you can adapt and combine over time. Instead of waiting for a perfect product or paying high consulting fees, you could start small, using proven APIs and existing automation platforms to create “good enough” solutions that save time and reduce errors.

Here’s how this can be done with the tools I work with daily. Using n8n, you can create workflows that connect Tripletex or Fiken to Vipps and NAV APIs, automating invoice generation, payment tracking, and compliance reporting. Integrating Telegram bots allows instant notifications and data collection without extra apps. For communication, simple terminal-based interfaces or light wrappers around messaging platforms can streamline customer or employee contacts. AI APIs like Azure OpenAI or Claude can assist in generating summaries or automating responses. Importantly, these solutions require no heavy programming or training custom ML models — just configuration and modular integration, which I specialize in.

This approach fits small to medium Norwegian companies ready to experiment and gradually improve their workflows without breaking the bank. It’s not a good fit for enterprises needing fully custom, large-scale systems or companies without any appetite for tech tinkering. The focus is on practical automation that improves daily operations and frees up human time.

So, what if instead of waiting for the perfect automation tool to appear, you started assembling your own set of smart helpers using existing APIs and low-code workflows? How much time and money could you save by taking small steps toward automation today?

Resources: github.com