1. Habit
  2. Skills
  3. AXI skill
kunchenguid · 5K installs

AXI skill
Runs by itself, CLI code changes.

Agent-facing CLIs stay AXI-compliant.

Install the AXI skill and make it a habit: follow https://gethabit.dev/skills/kunchenguid/axi.md

  1. 1Copy
  2. 2Paste into Claude Code, Codex or Cursor
  3. 3It runs by itself

5K installs2K starsMIT licenceSource on GitHub ↗

AXI

by kunchenguid · 5K installs

The skill

Axi defines Agent eXperience Interface standards for CLI tools agents use through the shell: TOON output, minimal schemas, truncation, structured errors, exit codes, session hooks and fast version checks.

Only when you ask.

The habit

Agent-facing CLIs stay AXI-compliant.

↻ Runs by itselfCLI code changes

How the habit works.
You don't have to.

Three questionsWhen does it start? How does it check its work? What does it remember?

01

It starts by itself

CLI code changes. Runs in GitHub Actions or Claude Code routine.

  • ↻ CLI code changes
  • pull_request paths: src/**, bin/**

02

It checks its work

flag and exit checks. If a check fails, it tells you.

  • ✓an unknown flag exits with code 2 and lists valid flags
  • ✓--version, -v and -V print the bare version and exit 0
  • ✓running the CLI with no arguments prints live content, not usage
  • ✓the version fast-path test passes

03

It remembers

AXI violation log. The next run reads it first: open AXI violations per command.

loops/agent-cli-ergonomics-loop/log.md

One run, step by step.
Then it stops.

Stop ruleStops after the changed commands of one pull request, at most 20 commands.

  1. 01

    Read the TOON spec

    Read the TOON specification before judging output.

  2. 02

    Check output

    Check TOON on stdout, 3-4 default fields, truncation with total size and total counts in lists.

  3. 03

    Check empty states

    Make sure empty results state the zero with context.

  4. 04

    Check errors

    Make sure errors go to stdout with a fix hint, mutations are idempotent and no prompts appear.

  5. 05

    Check flags and help

    Run an unknown flag, --help and --version and check exit codes and output.

  6. 06

    Log violations

    Record each violation with the command and the AXI section it breaks.

What it may do.
Nothing more.

Safe by defaultIt writes only in its own folder and asks you before anything goes out.

Is it for you?
Use it when it fits.

Quick checkRead both lists. If the left one sounds like you, the AXI skill helps.

Use it when

  • ✓ You build a CLI for agents
  • ✓ You review CLI output for token cost
  • ✓ You add session hooks to a tool

Skip it when

  • – Your CLI is only for humans
  • – You build a GUI app

Questions

AXI skill, answered.

Why reject unknown flags instead of ignoring them?

The skill says a silently dropped flag gives the agent plausible but wrong output.

Where do progress messages go?

To stderr; stdout carries only structured output the agent reads.

How do I install the AXI skill?

Copy the prompt at the top into your agent. It installs the skill from kunchenguid/axi and sets up the habit. Or run npx skills add kunchenguid/axi --skill axi.

Is the AXI skill free?

Yes. It is published under MIT. Habit is free too.

AXI. Now as a habit.

Install the AXI skill and make it a habit: follow https://gethabit.dev/skills/kunchenguid/axi.md

  • Free
  • No account
  • Undo in one line

All kunchenguid skills →More code skills →

Works in

  • Claude Code
  • Codex
  • Cursor
  • ChatGPT
  • Claude app

Made by kunchenguid, published under MIT. Habit links to the original and never changes it. Is this your skill?

Copying didn't work here. Select the line and copy it by hand.