Token导航 LogoToken导航TokenDH.com
开发需要联网github未标认证来源可访问许可证需确认审计通过

document-feature文档特征

Agent Skill

document-feature 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。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

396

周安装

16

GitHub Stars

公开资料未说明

下载量

124
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:document-feature(文档特征)
来源仓库:https://github.com/folio-org/folio-eureka-ai-dev
仓库路径:skills/document-feature
安装命令:
npx skills add https://github.com/folio-org/folio-eureka-ai-dev --skill document-feature
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/folio-org/folio-eureka-ai-dev --skill document-feature

简介

该技能分析后端功能变更并生成行为级文档,聚焦 Spring/Quarkus 服务。

  • 适用于 FOLIO 类图书馆系统或类似 Java 后端项目的 API 文档维护。
  • 结合 OpenAPI YAML 与代码变更自动推导接口契约与异常场景。
  • 安装前请确认项目符合预设假设,否则需手动补充缺失信息。
  • document-feature 属于开发类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

Document Feature

You are a documentation specialist. After implementing a feature, analyze the code changes and generate behavioral feature documentation.

Scope and assumptions (this skill is intentionally scoped)

This skill targets backend modules with these common characteristics:

  • Java service built on Spring (Spring MVC) or Quarkus (JAX-RS)
  • PostgreSQL persistence
  • REST API documented in OpenAPI YAML (preferred source of truth for endpoints)
  • Kafka used for message consumption/production
  • Deployed as multiple instances behind a load balancer (clustered)

If a repo deviates, proceed best-effort but follow the evidence rules.

Non-negotiables

Feature = observable behavior

A feature is what external consumers can observe or interact with. Implementation mechanisms are aspects of features.

Feature (behavior)Aspect (mechanism)
Resource lookup APICache for lookup
Validation rulesException mapping
Event-driven state syncKafka listener wiring

Name features after behavior, not the mechanism:

  • Good: resource-lookup, hold-request-validation, tenant-sync-processing
  • Bad: resource-cache, validation-refactor, kafka-handler

Evidence-only rule (no guessing)

Only document endpoints/topics/config/integrations that you can point to in repo evidence:

  • OpenAPI spec (*.yml/*.yaml containing openapi: or swagger:)
  • Application config (application.yml, application.yaml, application.properties)
  • Code evidence (annotations, constants, clearly named configuration keys)
  • README or other checked-in docs

If you cannot prove it, omit it. Do not infer "likely" dependencies.

Questions

  • Write/update docs immediately.
  • Ask exactly one targeted question only when the feature boundary/name is genuinely ambiguous.
  • If changes are clearly refactoring/formatting/tests-only with no observable behavior change: stop and report that no feature doc update is needed.

Outputs

For each feature affected:

  • docs/features/<feature_id>.md (create directories if missing)
  • docs/features.md index (create if missing; if present, update minimally in existing style)

Feature doc frontmatter

Feature docs must include exactly these required frontmatter fields:

  • feature_id: must equal the file name (without .md)
  • title: human-readable Title Case
  • updated: YYYY-MM-DD (use today)

Existing docs may have extra frontmatter keys; preserve them. Do not add new keys (for example, do not introduce owners).

If an existing doc's feature_id does not match the filename: update feature_id to match the filename (do not rename files).

Documentation structure (fixed order; omit non-applicable sections)

Each feature document lives at docs/features/<feature_id>.md using this template.

---
feature_id: <kebab-case-id>
title: <Human Title>
updated: <YYYY-MM-DD>
---

# <Human Title>

## What it does
<2-3 sentences describing observable behavior from an external perspective.>

## Why it exists
<Business rationale: what problem it solves and why this behavior matters.>

## Entry point(s)
<Choose the relevant entry point representation(s). Omit only if the feature has no clear entry point.>

## Business rules and constraints
- <Rule 1 in plain language>
- <Rule 2>

## Error behavior (if applicable)
- <Externally visible error conditions and outcomes: status codes, validation failures, retry/idempotency expectations.>

## Caching (if applicable)
<Document caching only when it affects externally observable behavior or operational correctness in a cluster.>

## Configuration (if applicable)
| Variable | Purpose |
|----------|---------|
| <property.key or ENV_VAR> | <What it controls in this feature> |

## Dependencies and interactions (if applicable)
<Only feature-relevant, evidenced external interactions.>

For entry point format templates (REST, Kafka Consumer, Scheduled Job, Internal Event), see references/entry-point-templates.md.

The index file (docs/features.md)

If docs/features.md does not exist, create a minimal index:

# Module Features

This module provides the following features:

| Feature | Description |
|---------|-------------|
| [<Human Title>](features/<feature_id>.md) | <One-sentence behavioral description> |

If it exists but uses a different format, update minimally in the existing style (do not rewrite/normalize).

For detailed workflow steps (preflight, feature identification, entry points, behavior extraction, configuration, dependencies), see references/workflow.md.

Quick sanity checks

  • Feature names reflect behavior (not caching/events/implementation).
  • Every endpoint/topic/config/integration mentioned is backed by evidence.
  • Sections are in fixed order; non-applicable sections are omitted.

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

38.08%
按下载量换算47

Claude

31.31%
按下载量换算39

Cursor

18.77%
按下载量换算23

Gemini CLI

9.14%
按下载量换算11

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

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