.claude/skills/spec-create/SKILL.md
Create a new feature specification and update progress.md and roadmap.md
npx skillsauth add peintangos/deep-agents-example spec-createInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Arguments: $ARGUMENTS (PRD slug and feature name)
$ARGUMENTS, use itdocs/prds/*/prd.md using Glob
The target PRD path is docs/prds/prd-{slug}/ below.
docs/prds/prd-{slug}/prd.md and identify which PRD requirement this feature corresponds to. Also read docs/ubiquitous/glossary.md to use consistent terminology in the specificationdocs/prds/prd-{slug}/specifications/ for duplicates. Check existing spec numbers and determine the next numberdocs/prds/prd-{slug}/specifications/spec-NNN-<slug>.md using docs/prds/_template/specifications/spec-001-example.md as the baselinedocs/prds/prd-{slug}/progress.md using the exact template schema from docs/prds/_template/progress.mddocs/prds/prd-{slug}/dependencies.mddocs/roadmap.md as needed# spec-NNN: [Feature Name]
## Overview
[What this specification covers]
## Acceptance Criteria
```gherkin
Feature: [Feature Name]
Background:
[Background description for this feature]
Scenario: [Scenario name 1]
Given [Precondition]
When [Action/Event]
Then [Expected result]
Scenario: [Scenario name 2]
Given [Precondition]
When [Action/Event]
Then [Expected result]
```
## Implementation Steps
- [ ] [First executable task]
- [ ] [Next executable task]
- [ ] [Add or update tests]
- [ ] Review (build check + lint + `/code-review`)
## Specification StatusSpecification | Title | Status | Completed On | Notesspec-NNN-<slug>.md filepending as the initial statusCompleted On and Notes empty for a new specification## Summary so the done count and current focus remain accuratedocs/prds/prd-{slug}/todo.md- [ ] spec-NNN: [task summary]/implement run or one Ralph iterationspec-NNN-<slug>.md (e.g., spec-001-transaction-crud.md)progress.md must keep one row per specification file; do not invent alternate table layoutstodo.md executable tasks must remain unchecked checkbox lines (- [ ])/code-review) at the end of implementation stepsdocs/ubiquitous/glossary.mddata-ai
監査レポートを Zenn 技術記事向けの文体 (だ/である調 + 比較表 + 階層化された見出し) に整えるスキル。`src/reporter.ts` の骨組みを LLM が整形してレポート (`out/*.md`) を仕上げるときに読み込む。
testing
OSS リポジトリの既知脆弱性 (OSV / GHSA) を照合し、重大度と影響範囲を分類するスキル。セキュリティ監査を実行するときに読み込む。
data-ai
OSS リポジトリのメンテナンス健全性 (リリース頻度・Issue 対応速度・放置 PR) を定量的に評価するスキル。メンテナンス監査を実行するときに読み込む。
tools
OSS リポジトリのメインライセンスを特定し、商用利用制約・依存互換性・NG ライセンスの検出までを行うスキル。ライセンス監査を実行するときに読み込む。