本文へ移動
cccskills
無料GitHub で公開

ase-meta-mint

Mint an identifier or a name of a requested type out of a free-text hint: a UUID, a SHA-1 digest, a constant, variable, function, or class identifier, a path component, or a product-like name. Use when the user wants an identifier, id, symbol, slug, or name "minted", "coined", "derived", or "generated" from a description.

インストール方法を見る

含まれるファイル(2)

  • SKILL.md5.7 KB
  • help.md3.7 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

@${CLAUDE_SKILL_DIR}/../../meta/ase-control.md @${CLAUDE_SKILL_DIR}/../../meta/ase-skill.md @${CLAUDE_SKILL_DIR}/../../meta/ase-getopt.md

<purpose name="ase-meta-mint"> Mint an Identifier or Name </purpose>

<expand name="getopt" arg1="ase-meta-mint" arg2="--type|-t=(uuid|sha1|const|var|class|func|path|name) --count|-c=1"> $ARGUMENTS </expand>

<objective> *Mint* an identifier or a name of the requested type out of the following hint: <hint><getopt-arguments/></hint> </objective> <flow>
  1. <step id="STEP 1: Sanity Check Usage">
    1. Set <type><getopt-option-type/></type> and set <hint/> to <getopt-arguments/> with any leading and trailing whitespace stripped.

    2. Determine the number of identifiers to mint: set <count/> to <getopt-option-count/>; if <getopt-option-count/> is non-numeric, less than 1, or greater than 100, use the default 1 instead.

    3. <if condition="<hint/> is empty AND <type/> is NOT `uuid`"> Only output the following <template/> and then immediately *STOP* processing the entire current skill -- every type except `uuid` needs a hint to derive anything from: <template> ⧉ **ASE**: ✪ skill: **ase-meta-mint**, ▶ ERROR: expected a `<hint>` argument for type **<type/>** </template> </if>
    4. Set <results></results> and <warning></warning> (set both to empty).

    </step>
  2. <step id="STEP 2: Mint Hash-Derived Identifiers" condition="<type/> is `uuid` or `sha1`">
    1. Call the ase_mint(type: "<type/>", hint: "<hint/>", count: <count/>) tool of the ase MCP server and set <results/> to the returned lines of its text output, one identifier per line. You MUST NOT derive a UUID or a SHA-1 digest yourself -- only the tool hashes correctly and reproducibly.

    2. <if condition="a line of the `text` output starts with `WARNING:`"> Remove that line from <results/> and set <warning/> to its text with the leading `WARNING:` stripped -- hashing a non-empty hint is deterministic, so <count/> was clamped to *1*. </if>
    </step>
  3. <step id="STEP 3: Derive Linguistic Identifiers" condition="<type/> is `const`, `var`, `class`, `func`, `path`, or `name`">
    1. Distill <hint/> into its essential words: drop filler words, articles, and prepositions, keep the two to four words which carry the actual meaning, and order them from the most general to the most specific one. Prefer established terminology of the surrounding domain over invented wording. Do not output anything.

    2. Assemble <count/> distinct candidates from these words, ordered best-first, honoring the rules of <type/>:

      • const: an UPPER_SNAKE_CASE constant identifier such as FOO_BAR_QUUX -- upper-case ASCII words joined by _, never starting with a digit.

      • var: a lowerCamelCase variable identifier such as fooBarQuux -- the last part MUST be a substantive (Quux) naming what the value is, never a verb, as a variable holds a thing and not an action.

      • func: a lowerCamelCase function identifier such as fooBarQuux -- the first part MUST be a verb (foo) naming what the function does, as a function performs an action and does not hold a thing.

      • class: an UpperCamelCase class identifier such as FooBarQuux -- a substantive phrase naming the modeled entity, in singular form and without a The prefix.

      • path: a kebab-case path component such as foo-bar-quux -- lower-case ASCII letters, digits, and - only, never starting or ending with -.

      • name: an UpperCamelCase product name such as FooBarQuux -- optimize for a brand, not for a description: short (preferably two to three syllables), pronounceable, memorable, and distinctive. Coinages, blends, and evocative metaphors are explicitly welcome here, and a plain concatenation of the hint words is explicitly not.

      Store the candidates in <results/>. Do not output anything.

    </step>
  4. <step id="STEP 4: Render Result">
    1. <if condition="<hint/> is empty"> Set <hint-shown>*(none -- random UUID)*</hint-shown>. </if> <else> Set <hint-shown><hint/></hint-shown>. </else>
    2. Output the minted identifiers of <results/>, one per line and each enclosed in backticks, with the following <template/>:

      <template> <ase-tpl-boxed title="MINT" subtitle="<type/>"> ● **HINT**: <hint-shown/>

      ○ <result-1/> ○ <result-2/> ○ [...] </ase-tpl-boxed> </template>

    3. <if condition="<warning/> is non-empty"> Additionally output the following <template/> directly below the box: <template> ⧉ **ASE**: ✪ skill: **ase-meta-mint**, ▶ WARNING: <warning/> </template> </if>
    </step>
</flow>

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Review software architecture, including package cohesion and inter-package coupling

日本語の概要は準備中です。原文の説明を表示しています。

rse/ase592026年10月11日 更新

Discover additional, third-party components (libraries/frameworks) for the technology stack to provide needed functionality.

日本語の概要は準備中です。原文の説明を表示しています。

rse/ase592026年10月11日 更新

Analyze the source code for problems in either the logic and semantics and its related control flow, performance and efficiency, or security.

日本語の概要は準備中です。原文の説明を表示しています。

rse/ase592026年10月11日 更新

Craft Source Code: Use when user wants to "create", "add", or "craft" a new feature from scratch.

日本語の概要は準備中です。原文の説明を表示しています。

rse/ase592026年10月11日 更新

Edit Source Code: Use when the user wants to "edit" the code base in one shot from a query or a bare analyzer issue id like "P1", fusing crafting, refactoring, and resolving with optional grilling, verification, looping, and Git worktree isolation.

日本語の概要は準備中です。原文の説明を表示しています。

rse/ase592026年10月11日 更新

Explains code with WHAT, WHY, ANALOGY, DIAGRAM, CRUXES, and GOTCHAS. Use when you want to know how code works or when the user asks "how does this work?"

日本語の概要は準備中です。原文の説明を表示しています。

rse/ase592026年10月11日 更新

rse のスキルをすべて見る

このスキルの問題を報告する