Token导航 LogoToken导航TokenDH.com
研究检索只读github未标认证来源可访问许可证需确认审计通过

canonical-spec-format规范规范格式

Agent Skill

canonical-spec-format 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

186

周安装

8

GitHub Stars

61

下载量

65
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:canonical-spec-format(规范规范格式)
来源仓库:https://github.com/melodic-software/claude-code-plugins
仓库路径:skills/canonical-spec-format
安装命令:
npx skills add https://github.com/melodic-software/claude-code-plugins --skill canonical-spec-format
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/melodic-software/claude-code-plugins --skill canonical-spec-format

简介

canonical-spec-format 提供跨供应商的统一规格格式参考,支持YAML/JSON结构定义。

  • 适用于API设计、接口文档编写及规范转换场景,兼容Codex、Claude等宿主。
  • 包含字段映射、验证规则和调试指引,帮助对齐不同服务商的实现差异。
  • 使用时需结合实际代码或样例,避免纯理论推导导致偏差。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

Canonical Specification Format

Reference guide for the canonical specification format - the provider-agnostic intermediate representation defined in ADR-115.

When to Use This Skill

Keywords: canonical specification, canonical spec, spec schema, specification format, provider-agnostic, spec fields, spec validation, spec structure, YAML specification, JSON schema

Use this skill when:

  • Understanding canonical specification structure
  • Validating specifications against schema
  • Creating specifications manually
  • Mapping between providers and canonical format
  • Debugging specification transformation issues

Quick Reference

Minimal Valid Specification

id: "SPEC-001"
title: "Feature Title"
type: feature

context:
  problem: "Problem description (min 20 chars)"
  motivation: "Business value"

requirements:
  - id: "REQ-001"
    text: "The system SHALL do something"
    priority: must
    ears_type: ubiquitous
    acceptance_criteria:
      - id: "AC-001"
        given: "precondition"
        when: "action"
        then: "outcome"

metadata:
  status: draft
  created: "2025-12-24"
  provider: canonical

Required Fields

FieldTypeDescription
idstringFormat: SPEC-{number}
titlestringHuman-readable title
typeenumfeature, bug, chore, spike, tech-debt
context.problemstringMin 20 characters
context.motivationstringBusiness value
requirementsarrayAt least one requirement
metadata.statusenumdraft, review, approved, implemented, deprecated
metadata.createdstringISO 8601 date
metadata.providerstringProvider that created this spec

Field Reference

Root Level

id: "SPEC-001"           # Required: Unique identifier
title: "Feature Title"    # Required: Human-readable name
type: feature             # Required: Specification type

Type Values:

TypeDescription
featureNew functionality or capability
bugDefect fix
choreMaintenance task
spikeResearch or investigation
tech-debtTechnical debt reduction

Context Section

context:
  problem: |                    # Required: min 20 chars
    Clear description of the problem.
    What pain point does this address?
  motivation: |                 # Required
    Business value or user benefit.
    Why should we invest in this?
  background: |                 # Optional
    Additional context, history, constraints

Requirements Section

requirements:
  - id: "REQ-001"               # Required: Unique within spec
    text: "EARS requirement"    # Required: EARS-formatted
    priority: must              # Required: must/should/could/wont
    ears_type: event-driven     # Required: EARS pattern type
    acceptance_criteria:        # Required: at least one
      - id: "AC-001"
        given: "precondition"
        when: "action"
        then: "outcome"
        and:                    # Optional: additional conditions
          - "additional condition"
    notes: "Optional notes"     # Optional

Priority Values (MoSCoW):

PriorityDescription
mustNon-negotiable, system fails without it
shouldImportant but not critical
couldDesirable if resources permit
wontExplicitly excluded from scope

EARS Type Values:

TypePatternExample
ubiquitousThe system SHALL..."The system SHALL encrypt data"
state-drivenWHILE..., the system SHALL..."WHILE active, the system SHALL..."
event-drivenWHEN..., the system SHALL..."WHEN clicked, the system SHALL..."
unwantedIF..., THEN the system SHALL..."IF error, THEN the system SHALL..."
complexCombinations"WHILE active, WHEN clicked..."
optionalWHERE..., the system SHALL..."WHERE enabled, the system SHALL..."

Design Section (Optional)

design:
  approach: |                   # Optional: implementation approach
    High-level description of how to implement
  components:                   # Optional: affected components
    - "Component 1"
    - "Component 2"
  dependencies:                 # Optional: prerequisites
    - "External dependency"
  alternatives:                 # Optional: considered alternatives
    - name: "Alternative approach"
      reason_rejected: "Why not chosen"

Traceability Section (Optional)

traceability:
  adr_refs:                     # Optional: related ADRs
    - "ADR-115"
  requirement_refs:             # Optional: related requirements
    - "FR-001"
    - "NFR-001"
  epic_ref: "EPIC-1118"         # Optional: parent EPIC
  user_story_refs:              # Optional: related user stories
    - "US-001"

Metadata Section

metadata:
  status: draft                 # Required
  created: "2025-12-24"         # Required: ISO 8601
  provider: canonical           # Required
  version: "1.0.0"              # Optional: semantic version
  bounded_context: "WorkManagement"  # Optional: from ADR-024

Status Values:

StatusDescription
draftInitial creation, not reviewed
reviewUnder review/refinement
approvedApproved for implementation
implementedImplementation complete
deprecatedNo longer valid

Bounded Context Values (ADR-024):

  • WorkManagement
  • Orchestration
  • Workflows
  • Expertise
  • ExecutionControl
  • TriggerManagement
  • Integrations

Validation Rules

ID Formats

FieldFormatExample
Specification IDSPEC-{number}SPEC-042
Requirement IDREQ-{number}REQ-001
Acceptance Criterion IDAC-{number}AC-001
ADR ReferenceADR-{number}ADR-115
EPIC ReferenceEPIC-{number}EPIC-1118
User Story ReferenceUS-{number}US-001

Content Constraints

FieldConstraint
context.problemMinimum 20 characters
requirementsAt least one requirement
acceptance_criteriaAt least one criterion per requirement
metadata.createdValid ISO 8601 date

EARS Pattern Validation

Each requirement's text field must match its declared ears_type:

ears_typeRequired Pattern
ubiquitousStarts with "The" + entity + "SHALL"
state-drivenStarts with "WHILE"
event-drivenStarts with "WHEN"
unwantedContains "IF" and "THEN"
optionalStarts with "WHERE"
complexMultiple pattern keywords

JSON Schema Location

schemas/canonical-spec.schema.json

Provider Transformation

The canonical format serves as the hub for all provider transformations:

                    ┌─────────────┐
                    │  Canonical  │
                    │    Spec     │
                    └──────┬──────┘
           ┌───────────────┼───────────────┐
           │       │       │       │       │
           ▼       ▼       ▼       ▼       ▼
        ┌─────┐ ┌─────┐ ┌─────┐ ┌─────┐ ┌─────┐
        │EARS │ │Ghrkn│ │Kiro │ │SpKit│ │ ADR │
        └─────┘ └─────┘ └─────┘ └─────┘ └─────┘

All providers implement ISpecificationProvider:

interface ISpecificationProvider
{
    string ProviderName { get; }
    Task<Result<CanonicalSpec>> ParseAsync(string input);
    Task<Result<string>> GenerateAsync(CanonicalSpec spec);
    Task<ValidationResult> ValidateAsync(CanonicalSpec spec);
    bool CanParse(string input);
}

References

Detailed Documentation:

Repository Resources:

  • schemas/canonical-spec.schema.json - JSON Schema
  • docs/adr/ADR-115-specification-provider-abstraction.md - Architecture decision

Last Updated: 2025-12-26

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

34.79%
按下载量换算23

Claude

29.96%
按下载量换算19

Cursor

21.31%
按下载量换算14

Gemini CLI

9.8%
按下载量换算6

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills