One CLI Blog
Engineering notes and product design records
Notes on One CLI manifests, presets, AI project guidance, and template governance so decisions stay traceable.
Agentic Engineering: Structured Workspaces for AI Coding Agents
Agentic engineering makes AI coding agents useful by giving them clear workspace state, repeatable commands, and fewer places to guess.
What an AI-Native Monorepo CLI Needs to Control
An AI-native monorepo CLI should give agents a stable workspace contract, not just a folder generator.
How Coding Agents Should Prepare a Workspace to Run
A safe workspace setup flow starts from the manifest, installs only missing dependencies, and reports exact commands.
Why CLI JSON Output Matters for Coding Agents
Stable JSON envelopes let agents branch on error codes and context instead of parsing human help text.
Why Monorepo Scaffolding Needs a Manifest
A monorepo scaffold should record why projects exist, not just where files were generated.
One CLI vs General Scaffolding Tools
One CLI focuses on workspace contracts and agent-safe operations, not only initial project generation.
Template Governance for AI-Ready Workspaces
Templates are safer for agents when they carry conventions, dependency rules, and generated guidance together.
Historical Design: One CLI Agent Skills
Historical notes on the agent skills design, retired when One CLI removed skill distribution and generated instructions.
Why One CLI Treats the Manifest as the Contract
one.manifest.json is not just scaffold output. It is the shared contract read by humans, the CLI, and coding agents.
What Preset IDs Should and Should Not Do
A preset ID is a reproducible template-combination code. It should not grow into another business-scenario DSL.