Consulting, Teaching and Mentoring in the Craft of Software Development |
Thinking Craftsman AI Toolkit (TC AI Toolkit)
Hosted on Github At : https://github.com/nitinbhide/tcaitoolkit
Recently I am working on helping software development teams adopt Agentic Engineering. This approach is very different from the casual "vibe coding" trend. Agentic Engineering is about using coding agents deliberately and effectively, with the goal of reducing the number of iterations required to fix a bug or implement a feature.
When teams adopt "agentic engineering" , everything improves—developer productivity rises, bugs and regressions decrease, and token usage becomes far more efficient. It's genuinely a win-win for everyone involved. But like any meaningful craft, it demands discipline.
I started teaching these skills and practices to teams across several companies. Through those experiences, I realized there was a need for a structured toolkit that supports this disciplined approach. That led me to create the Thinking Craftsman AI Toolkit, which is now available on GitHub
Following features are now available in Thinking Craftsman AI Toolkit
- Cookie Cutter to Generate a new software development project skeleton /folder structure. This is optimized for use/adoption of Coding Agents.
- 4 Different Agent Skills
- Thinking Craftsman Code Review Skill - for code review and new code generation
- Version Control Skill to use different version control (git, mercurial, subversion) with Coding Agents
- Skill to evaluate the Agent Skills
- A Grill me Skill adopted from 'Grill me' skill of mattpocock
The toolkit is released under the Apache License. I encourage you to explore it, experiment with it, and share your feedback—whether it’s suggestions for improvements or reports of bugs.