Design agent-first CLIs that emit JSON envelopes with HATEOAS-driven command trees.
npx skills add https://github.com/godaddy/cli --skill cli-design-godaddy
Please help me install this Agent Skill. Skill: cli-design Source: https://github.com/godaddy/cli/tree/main/.agents/skills/cli-design Command: npx skills add https://github.com/godaddy/cli --skill cli-design-godaddy
Command-line tools often expose textual output that is hard for agents to interpret and for humans to verify. This guide defines patterns for agent-first CLIs that always emit JSON envelopes, expose discoverable command trees via a root discovery, and protect context in outputs to maintain useful, concise results.
Install the CLI and run the root command to see the full command tree and example JSON envelopes.
High-intent search queries and answers about installing and using this skill.
An agent-first CLI is a command-line tool designed to emit JSON-only output, enabling reliable agent parsing and orchestration. It uses JSON envelopes to deliver self-describing command trees, protecting context and ensuring predictable, navigable responses for automation workflows.
To design a CLI with HATEOAS-driven navigation, implement a root discovery command that outputs a self-documenting command tree. This exposes prefilled next_actions within JSON envelopes, allowing agents to autonomously discover and execute subsequent steps without relying on external documentation.
You need streaming NDJSON output in a command-line tool for long-running operations. It maintains context-protecting output by streaming structured data incrementally, preventing context window overflow while providing agents with real-time, parseable JSON envelopes during extended tasks.
This approach works for both building new CLI tools and enhancing existing ones. It targets developer teams needing agent-driven workflows, allowing you to retrofit current tools to support JSON-only output, self-describing command trees, and predictable integration with automation systems.
A CLI should handle errors by returning consistent error envelopes within its JSON output. These envelopes include actionable fixes, ensuring that agents can programmatically interpret the failure, understand the required correction, and attempt a resolution without human intervention.
You should replace text-based CLI output with JSON envelopes because raw text is hard for agents to interpret and for humans to verify. JSON-only responses enable reliable agent parsing, while context-protecting output maintains useful, concise results for orchestration.
aleph23Automate browser tasks and debugging with Puppeteer-based CLI scripts.
Yogi-076Automate multi-step workflows with human approval checkpoints and JSON status envelopes.
wuyyyyyouStreams new Gmail messages as NDJSON output via configured sink.
joelhooksDesign agent-first CLIs with HATEOAS JSON responses and self-documenting command trees.
