本文へ移動
cccskills
無料GitHub で公開

eli5

Explain a technical spec or proposed change in plain language: what problem it solves, how the solution works, and every schema change.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md3.2 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。

ELI5

Turn the referenced spec or change into an accurate mental model for a technical reader who does not live in the code. Plain language, full precision: simplify the telling, never the claims.

The Register

Other skills point here when their output must be readable by someone who didn't live the work. The reader is technical — pseudocode and precise claims land fine — but they have read none of the code and none of the session's messages: no diff, no transcript, no labels the work invented along the way. Writing ELI5 means:

  • Walk one concrete scenario end to end — the triggering event, what happens today, what the change (or the unbuilt alternative) would do — instead of describing properties in the abstract.
  • Define every term of art at first use; never lean on labels the spec, code, or session invented.
  • Use small concrete examples without replacing precise claims with analogies.
  • Reach for pseudocode when explaining control flow, ordering, or timing. Prose describing when-something-fires reads as plausible and hides the gap; five lines of pseudocode make the gate visible and let the reader see the case you missed. Write it at the level of the decision, not the implementation — the conditions and their order, not real function signatures. The tell that you needed it: your prose contains "only when", "before", "unless", or "as soon as" and the reader still cannot say what happens on the second call.
  • The test: the text stands alone, without the diff, the spec, or the transcript. If the reader must ask "explain this part", it failed.

Workflow

  1. Read the complete referenced artifact and inspect current owners when the spec alone cannot establish behavior. When no artifact is named, explain the session's current change (working tree or branch diff). Separate what exists today from what is only proposed.
  2. Explain the problem through its user or operational consequence, including why the current design produces it.
  3. Explain the solution as one simple before/after data flow. Introduce each component by responsibility, not by filename or internal symbol.
  4. Inventory schema and durable-contract changes exhaustively: added, changed, removed, reset, and deliberately unchanged. Include cursor or wire-format cutovers when they affect stored data or readers. Say explicitly when there are no schema changes.

Output

Use these headings in order:

  1. Problem
  2. Solution
  3. Schema changes

End with one short sentence stating what users should notice after the change.

Rules

  • Lead with behavior and boundaries; mention implementation names only when they clarify ownership or a contract.
  • Distinguish canonical records from derived indexes, caches, summaries, and presentation grouping.
  • Call out destructive resets, migration requirements, compatibility behavior, eventual consistency, and intentional data loss directly.
  • Do not omit a schema change because it is operational rather than user-visible.

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

Audit or rewrite AGENTS.md so it holds only lasting principles. Run it only when the user asks for it by name; never invoke it on your own.

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。

dzhng/skills1,0292026年10月6日 更新

Audit the choices an implementing agent made, not its diff — a pure decision audit that traces the session's history into a choices ledger, changes no code, and never blocks an unsupervised run. Working code still embeds architecture the user never chose; surface it because future work inherits it. Use when the user wants to review the decisions the AI made on their behalf, before merging or committing AI-implemented work, when integrating a delegated subagent's pass, or when a fix "works" but might be a point fix.

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。

dzhng/skills1,0292026年10月6日 更新

Audit and prioritize performance work through bounded-work and forward-progress checks. Use when asked to find performance issues, investigate freezes/500s/OOMs, review retries or polling for no-progress loops, rank a performance backlog, or implement the next simple performance fixes.

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。

dzhng/skills1,0292026年10月6日 更新

Audit whether tests earn their maintenance cost and which suite owns each contract. Use when pruning redundant tests, investigating implementation coupling or test-only production hooks, reviewing the value of proposed coverage, or auditing an entire subsystem. Use write-tests to implement the resulting test changes.

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。

dzhng/skills1,0292026年10月6日 更新

Run fast, progressive experiments to map tunable parameters and their effects. Use when optimizing code, prompts, configurations, or other artifacts against a goal or benchmark, or when a long research run needs a planned hypothesis queue, focused trials, and durable findings.

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。

dzhng/skills1,0292026年10月6日 更新

claude

無料

Use Claude Code as an independent `claude -p` subagent when the user explicitly asks for Claude, wants a second-agent opinion from Claude, or asks to delegate a well-scoped task to Claude. Supports selecting `--model` and thinking/effort level with defaults of `opus` and `high`.

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。

dzhng/skills1,0292026年10月6日 更新

dzhng のスキルをすべて見る

このスキルの問題を報告する