check-coverage

Agent Skills › ldayton/Dippy › check-coverage

check-coverage

GitHub

确保CLI处理程序及测试具备全面覆盖率。用于新增命令或审计现有处理器时，通过查阅文档、探索子命令、审查和补充测试用例，直至验证通过。

.claude/skills/check-coverage/SKILL.md ldayton/Dippy

Trigger Scenarios

添加新的CLI命令审计现有处理程序的测试覆盖率

Install

npx skills add ldayton/Dippy --skill check-coverage -g -y

More Options

Non-standard path

npx skills add https://github.com/ldayton/Dippy/tree/main/.claude/skills/check-coverage -g -y

Use without installing

npx skills use ldayton/Dippy@check-coverage

指定 Agent (Claude Code)

npx skills add ldayton/Dippy --skill check-coverage -a claude-code -g -y

安装 repo 全部 skill

npx skills add ldayton/Dippy --all -g -y

预览 repo 内 skill

npx skills add ldayton/Dippy --list

SKILL.md

Frontmatter

{
    "name": "check-coverage",
    "description": "Ensure comprehensive test coverage for a CLI handler. Use when adding a new command or auditing existing handler coverage.",
    "argument-hint": "<tool>",
    "disable-model-invocation": true
}

Ensure the handler and tests for $ARGUMENTS provide comprehensive coverage.

1. Gather Documentation

Find documentation for the tool (need at least one source):

tldr:

ls ~/source/tldr/pages/*/$ARGUMENTS*.md
cat ~/source/tldr/pages/*/$ARGUMENTS.md

Local CLI:

$ARGUMENTS --help
man $ARGUMENTS

Stop if neither source exists.

2. Explore Subcommands

For tools with subcommands, recursively explore:

$ARGUMENTS <subcommand> --help
$ARGUMENTS help <subcommand>

Build a mental model of:

All subcommands and their actions
Which operations are read-only (safe) vs mutate state (unsafe)
Global flags that affect parsing
Edge cases

3. Review Existing Tests

Read tests/cli/test_$ARGUMENTS.py and check for:

Coverage of all subcommands
Both safe and unsafe variants of each action
Global flag handling
Edge cases from the docs

4. Add Missing Tests

Add aspirational test cases for anything missing. Follow existing format:

TESTS = [
    # --- Subcommand group ---
    ("$ARGUMENTS <subcommand> <safe-action>", True),
    ("$ARGUMENTS <subcommand> <unsafe-action>", False),
]