What Is Mochi Sticky?

Long story short: Mochi Sticky is a productivity tool for developers.

Long story long (because yes, this is a blog): it is a small tool I built to reduce context switching when I am juggling multiple tasks and projects.

Who am I?

Before I start, a quick introduction so you know who is talking.

I am a software developer and I have been working in the industry for a few years. I am genuinely happy I chose this profession, and I really enjoy building different solutions. If you want to know more about me, you can follow this link: About me (yes, I am slightly narcissistic).

We are living in the era where AI is disrupting how we develop software. I started using it in my day-to-day work, just like many of my coworkers.

The big difference between AI and normal tools/frameworks: most tools come with a manual, AI often starts with "good luck." So we do the classic trial-and-error dance and try not to spend all our credits.

How was Mochi Sticky born?

Across different projects I kept running into the same problem: context switching.

• I am working on application A, I get blocked, then I pick up another story that needs changes in application B.
• I am working on an application and suddenly a production fix takes priority.
• I am working on a service that requires coordinated changes across multiple microservices.

At work we already use tools to track stories/tasks, but I also wanted a place for baby steps, quick ideas, and must-have vs nice-to-have notes.

I tried different tools, but I kept returning to a simple txt file. Later it evolved to Markdown. I kept it inside the project, ignored it in Git, and used it as my personal brain RAM.

For a long time I had an idea spinning in my head: what if that Markdown workflow was an actual tool?

Then AI entered the chat. I tried different prompts and agent setups. Some produced great results; others were easy to drop and restart. My conclusion: AI agents are very good at small, concrete tasks.

After enough trial and error, I ended up doing a tiny SDLC with AI: Analysis -> Design -> Planning (break into small tasks) -> Implementation -> Test.

Around the same time, I came across Spec-Driven Development and tried GitHub Spec Kit (Spec-Driven Development). It matched how I wanted to work with agents, but sometimes I felt I was overcomplicating things, so I went back to my old approach.

Even with that failed attempt, I felt I was still on the right path: break work down, keep tasks concrete, keep the feedback loop tight.

Without noticing, I ended up with task lists spread across different Markdown files in different projects. That is when I decided to cook Mochi Sticky.

Mochi Sticky principles

• Keep information in Markdown, so I can read it in any IDE or in the terminal (cat, less, vi).
• Easy to use and easy to understand.
• Configurable, but not required to get started.
• Minimal dependencies: install it and use it.
• AI-friendly.

Mochi Sticky and AI

We are in the AI hype era, and everything gets an AI sticker for marketing. That is not what Mochi Sticky is.

Mochi Sticky does not have AI built in. It is a simple tool. But I did build it with AI agents, and at some points I was using Mochi Sticky while building Mochi Sticky.

Is this vibe coding? Maybe. I did not build it to say "I used vibe coding." I built it because I wanted the tool.

Why did I use AI?

I could say I mastered AI agents, but honestly: I am lazy. I love programming, but this is a side project and I do not have unlimited time.

I also wanted to compare tools. I started paying for ChatGPT for my family, and I realized I was not really using it for coding while at work we use other agents.

Why is it beta?

I think it already has enough functionality to be useful. But I want to polish a few things before a final release.

(Remember when Gmail was beta for years? I will try to do better.)

What is missing for me:

• Improve the TUI interface and navigability.
• A deeper code review/refactor.
• Add more security restrictions.

Conclusion

Mochi Sticky is a simple productivity tool (one of many). It is not groundbreaking, but I am happy with the product and the journey of making it.

If I release it as open source, I can check "Contribute to an open-source project" off my bucket list.

I do not have social media. I will keep posting progress and releases on this page. If the project gets more traction, I will consider another update channel.

Regards!

Disclaimer: I am not a native English speaker. This page was written by me, then corrected and edited by AI.