Skip to content
Skills Library
Programming · Review & audit

cli-design

/cli-design

AI 友善的 CLI 設計與稽核(判準取自 Cloudflare `cf` CLI)。 Use when 新寫 CLI、新增或改動命令與旗標、稽核既有 CLI 對 agent 友不友善。 NOT for CLI 壞掉找原因(diagnose)、shell script 的 sudo/trap 安全、只是用某支 CLI 完成任務。

Licence
MIT
Last repository change
27.10.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
9
Scripts
1
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 27.10.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What to check before installing
  • 1 scripts in this package. Review them before installing, as Claude may execute them.

Scripts: scripts/probe_cli.py

What it does and when to use it

AI 友善的 CLI 設計與稽核(判準取自 Cloudflare `cf` CLI)。 Use when 新寫 CLI、新增或改動命令與旗標、稽核既有 CLI 對 agent 友不友善。 NOT for CLI 壞掉找原因(diagnose)、shell script 的 sudo/trap 安全、只是用某支 CLI 完成任務。

Original author description

AI 友善的 CLI 設計與稽核(判準取自 Cloudflare `cf` CLI)。 Use when 新寫 CLI、新增或改動命令與旗標、稽核既有 CLI 對 agent 友不友善。 NOT for CLI 壞掉找原因(diagnose)、shell script 的 sudo/trap 安全、只是用某支 CLI 完成任務。

How it works

  1. Phase 1 -- 收斂對象與 lane
  2. Phase 2 -- 產出命令介面規格(design lane)
  3. Phase 3 -- 蒐集稽核證據(audit lane)
  4. Phase 4 -- 判定並產出稽核報告(audit lane)
  5. Phase 5 -- 驗證產出
Original SKILL.md headings · 34 lines, 160 words
SKILL.md preview
---
name: cli-design
description: >-
  AI 友善的 CLI 設計與稽核(判準取自 Cloudflare `cf` CLI)。
  Use when 新寫 CLI、新增或改動命令與旗標、稽核既有 CLI 對 agent 友不友善。
  NOT for CLI 壞掉找原因(diagnose)、shell script 的 sudo/trap 安全、只是用某支 CLI 完成任務。
metadata:
  author: clade
  version: "1.0"
---


# CLI Design

設計與稽核共用同一份判準:每一項有 ID,設計時逐項滿足,稽核時逐項判定。下文 `<skill 根目錄>` 指本 skill 所在目錄(Claude Code 載入時印出的 `Base directory for this skill`)。

# SOP

## Phase 1 -- 收斂對象與 lane

1. READ 讀取使用者需求與目標 CLI 的原始碼入口、文件與既有 `--help`,判定走 design lane(新 CLI,或新增、改動命令介面)還是 audit lane(稽核既有 CLI),並確認主要使用者(agent、CI、人)與規模(單支 script 或多層命令樹)。
2. READ 讀取 `rules/AI友善CLI設計判準.md`,依其適用條件與已確認的使用者、規模,列出本次適用的判準 ID;design lane 進 Phase 2,audit lane 進 Phase 3。

## Phase 2 -- 產出命令介面規格(design lane)

1. THINK 依已載入判準逐項 ID 設計命令樹、共通旗標、輸出契約、錯誤與退出碼、非互動行為、破壞性操作、可發現性、認證設定與一致性治理;與既有命令衝突時,以已載入判準的詞彙為準並列出遷移方式。
2. WRITE 先讀取 `templates/cli-design-spec.md` 與 `templates/cli-design-spec.example.md`,再依骨架複製結構、參考範例改寫內容,產出規格寫入使用者指定位置(未指定時寫進 scratchpad)。
3. WRITE 若使用者要求實作,依規格實作,完成後以實作出的 CLI 為對象進 Phase 3;未要求實作則進 Phase 5。

## Phase 3 -- 蒐集稽核證據(audit lane)

1. READ 讀取 `rules/稽核探測與判定判準.md`,從原始碼確認 `--help` 會被攔截,以及目標 CLI 不帶引數與遇到看不懂的輸入時做什麼,再依已載入規則決定本次 `--allow`、`--typo` 與 `--depth`。
2. DELEGATE 執行 `uv run <skill 根目錄>/scripts/probe_cli.py [--allow <已確認安全的探測>] [--typo <名稱>] [--depth <N>] -- <CLI 命令與前綴參數>`,取得 help、版本、未知命令、未知旗標、色碼、等待輸入、子命令與旗標清單的 JSON 事實。
3. READ 讀取目標 CLI 原始碼與文件,補齊探測拿不到的判準項(破壞性操作的確認路徑、退出碼對照、錯誤輸出串流、認證優先序)。

## Phase 4 -- 判定並產出稽核報告(audit lane)

1. THINK 依已載入規則逐項 ID 判定 PASS、PARTIAL、FAIL 或 N/A,每格對應一筆探測輸出或原始碼位置,FAIL 與 PARTIAL 寫出修法,再依嚴重度排序。
2. WRITE 先讀取 `templates/cli-audit-report.md` 與 `templates/cli-audit-report.example.md`,再依骨架複製結構、參考範例改寫內容,產出報告寫入使用者指定位置(未指定時寫進 scratchpad)。

## Phase 5 -- 驗證產出

1. READ 回頭檢查產出是否符合已載入規則:規格的每個適用 ID 都有對應設計或 N/A 理由;報告的每格判定都有證據、每個 FAIL 都有修法、未探測項都寫了原因;不符合就回到對應 phase 修正。
Files in this package (9)
  • SKILL.md 3 KB
  • references/來源摘錄.md 5 KB
  • rules/AI友善CLI設計判準.md 20 KB
  • rules/稽核探測與判定判準.md 8 KB
  • scripts/probe_cli.py 11 KB
  • templates/cli-audit-report.example.md 11 KB
  • templates/cli-audit-report.md 2 KB
  • templates/cli-design-spec.example.md 8 KB
  • templates/cli-design-spec.md 2 KB

Versions

RepositoryScoreLicence
This page · YuDefine/nuxt-supabase-starter100MIT
caarlos0/dotfiles100MIT
CaptainCrouton89/crouton-kit100GPL-3.0
citypaul/.dotfiles95MIT
bengous/claude-code-plugins90MIT

Database snapshot: 13.11.2026 · package version c072d5bbc7fb (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 96tm-cli
    Use if and only if you use tm-cli (the Telemetry Machine CLI).
  • 100cloudflare-cf-cli
    Operates Cloudflare's new unified `cf` CLI (technical preview, April 2026) — install path, flag conventions, the local-vs-remote default trap, coexistence with Wrangler and `wrangler.jsonc`, and agent-mode usage via the Local Explorer OpenAPI. Use when the user mentions `cf`,…
  • 100cf-cli
    Cloudflare's official cf CLI (beta) for the public Cloudflare API and Workers projects. Use for finding and running Cloudflare commands from the terminal — sign-in and profiles, cf cli search, cf schema, cf deploy/dev/build, resource commands for DNS, zones, D1, KV, and R2 — and…
  • 100cli-agent-friendly-audit
    Audit a CLI tool against the agent-friendliness checklist from Zbigniew Sobiecki's "Building Agent-Friendly CLIs". Use when the user says "quality-cli-agent-friendly-audit", "cli-agent-friendly-audit", "audit this cli", "is my cli agent-friendly", or asks to check a CLI for agent ergonomics.
  • 100agent-native-design
    Use when designing, reviewing, or refactoring a CLI that must serve AI agents alongside humans, or when converting an API or SDK into an agent-usable CLI interface.
  • 100create-cli
    This skill should be used when the user asks to "design a CLI", "create command-line interface", "design CLI parameters", "plan CLI UX", or needs help structuring command-line tool interfaces with proper arguments, flags, subcommands, and user experience patterns.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.