PUKU CLI is a terminal-based AI coding tool built for Poridhi.io learners. It supports end-to-end development workflows directly from the command line: planning changes, editing and refactoring code, running tests, and iterating based on results—without switching between separate chat tools and scripts.

The tool is designed around task-driven agent workflows. Given a goal (for example, implementing a feature, fixing a bug, or preparing a release), PUKU CLI generates a plan, applies changes in a controlled sequence, runs verification steps, and surfaces the reasoning behind each decision. This makes it suitable for learning environments where users need both speed and transparency: the system helps, but it doesn’t hide the underlying engineering steps.

PUKU CLI also includes DevOps/SRE-oriented workflows as first-class capabilities. In practice, many learner projects involve containers, Kubernetes deployments, logs/metrics, and failure modes that look like production systems. The CLI provides structured routines for common operational tasks—debugging, environment validation, deployment checks, and automation—so learners practice the same loops used in real teams.

I was the Product Engineer for PUKU CLI. I designed the core interaction model and workflow architecture: how tasks are specified, how changes are executed and validated, how tool actions are sequenced, and how the experience stays consistent and approachable for learners while remaining grounded in real engineering practices.