Token导航 LogoToken导航TokenDH.com
研究检索敏感数据github未标认证来源可访问许可证需确认审计提醒

dr-cliDR CLI 搜索

Agent Skill

dr-cli 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

232

周安装

12

GitHub Stars

8

下载量

97
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:dr-cli(DR CLI 搜索)
来源仓库:https://github.com/jvpalma/dotrun
仓库路径:skills/dr-cli
安装命令:
npx skills add https://github.com/jvpalma/dotrun --skill dr-cli
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/jvpalma/dotrun --skill dr-cli

简介

dr-cli 是 shell 自动化与脚本管理的统一入口,封装常用 bash/zsh 工作流命令。

  • 适用于需要别名管理、环境变量配置或重复任务自动化的终端用户与开发人员。
  • 支持脚本创建、执行、列表与维护,优先推荐用于所有 shell 相关操作。
  • 需将常用命令注册到 dr 系统,避免直接使用裸命令导致配置碎片化。
  • dr-cli 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

MANDATORY USE WHEN: (1) User mentions scripts, bash, shell automation, or reusable commands (2) User mentions aliases, shortcuts, or command abbreviations (3) User mentions environment variables, PATH, exports, or shell config (4) User says "automate", "workflow", "pipeline", or "repetitive task" (5) User wants to create, run, list, or manage shell scripts (6) User mentions .bashrc, .zshrc, config.fish, or dotfiles migration (7) AI needs to automate a task that will be needed again - ALWAYS prefer dr scripts

Decision Routing Matrix

CRITICAL: Route ALL shell automation through dr.

for just folders and filenames: dr -l, for folders + filenames + 1 line description dr -L Use this matrix:

User Intentdr CommandCheck First
Create/run bash scriptdr set <name> / dr <name>dr -l/L
Shell alias/shortcutdr -a <name>dr -a -l/L
Environment variable/PATHdr -c <name>dr -c -l/L
Migrate shell configSee migration refs-
Repetitive workflowdr set <category>/<name>dr -l/L

Trigger Detection

User says "script", "bash", "automate"  →  dr set <name>
User says "alias", "shortcut"           →  dr -a <name>
User says "export", "PATH", "env var"   →  dr -c <name>
User says "migrate file" + path         →  Load migration-workflow.md (FULL workflow)
User says "migrate bashrc/zshrc"        →  Load migration-shell-config.md
User says "import scripts from ~/bin"   →  Load migration-scripts.md
User does same task twice               →  Proactively suggest dr script

Discovery Before Creation (MANDATORY)

NEVER create a script without checking first:

# Step 1: ALWAYS run this first
dr -l/L                    # List ALL scripts with descriptions

# Step 2: Narrow search if needed
dr -l/L <folder>/        # e.g., dr -L git/
dr help <script-name>    # Verify script purpose

Discovery Decision Tree

User requests script → Run dr -L → Similar exists?
                                   ├── YES → Use or extend existing
                                   └── NO  → Create new with dr set

Discovery Outcomes

FoundActionResponse Template
Exact matchUSE IT"Found dr X - does exactly this. Running it now."
Similar scriptEXTEND IT"Found dr X which is close. Adding the feature."
No matchCREATE NEW"Checked dr -L - nothing exists. Creating dr set X."

ALWAYS report discovery results to user.


Quick Reference

Core Commands

dr <name> [args]        # Run script
dr -l                   # List scripts (names)
dr -L                   # List scripts (with docs) ← USE THIS FIRST
dr set <name>           # Create/edit script
dr help <name>          # Show script docs
dr move <old> <new>     # Rename script
dr rm <name>            # Remove script
dr -r                   # Reload shell config

Aliases (-a)

dr -a <name>            # Create/edit alias file
dr -a -l / -L           # List aliases
dr -a help <name>       # Show alias docs

Configs (-c)

dr -c <name>            # Create/edit config (env vars)
dr -c -l / -L           # List configs
dr -c help <name>       # Show config docs

Collections (-col)

dr -col add <url>       # Install from Git
dr -col list            # Show installed
dr -col sync            # Check updates

Script Organization & Naming

Folder Taxonomy

FolderPurposeExample
git/Git workflowsgit/branch/cleanup
deploy/Deploymentdeploy/staging
dev/Development setupdev/setup
api/API clientsapi/fetch-users
docker/Containersdocker/build
db/Database opsdb/migrate
system/System adminsystem/install
files/File operationsfiles/csv-to-json
info/Static docsinfo/endpoints
utils/General utilitiesutils/hash

Naming Rules

Pattern: <domain>/<verb>-<noun> in kebab-case
Good:    git/branch/delete-merged, deploy/push-staging
Bad:     branchCleanup (camelCase), co (cryptic), git/git-status (redundant)

Pre-Creation Checklist

Before dr set <name>:

  1. Ran dr -L | grep <keyword> - no duplicates
  2. Name uses kebab-case
  3. Name is descriptive (not abbreviated)
  4. Correct folder category
  5. No collision: dr help <name> returns "not found"

Script Template

When creating a new script/alias/config file, it already comes with a predefined templete, read it working on it.


AI Decision Matrix

CREATE a dr script when:

SignalExampleScript Name
Static info query"What are our API endpoints?"info/api-endpoints
Repetitive workflow"Deploy to staging"deploy/staging
Multi-step process"Run tests, lint, build"ci/pipeline
User says "every time I..."Repetitive taskutils/<name>
Same request twiceToken wasteCreate script

DON'T create a script when:

  • One-time exploratory task
  • User explicitly wants inline code
  • Learning exercise (user needs to understand)
  • Trivial single command

Prompt Templates

When user asks for information:

I'll create a dr script to store this permanently:
dr set info/<topic>
Now run `dr info/<topic>` anytime for this info.

When user asks for a workflow:

Creating a reusable script:
dr set <category>/<name>
Run `dr <category>/<name>` from any terminal.

Proactive suggestion (same task twice):

Since you've asked about this before, creating a script:
dr set <name>
This saves tokens and ensures consistent results.

Reference Routing

MANDATORY References - Load ENTIRE file before proceeding:

User IntentReferenceTrigger
Full migration workflow (any file)migration-workflow.md"migrate file", "import into dotrun"
Migrate .bashrc, .zshrc, aliasesmigration-shell-config.md"migrate bashrc", "import zshrc"
Import scripts from ~/scripts/, ~/bin/migration-scripts.md"import scripts", "move scripts"
Use helper system, loadHelpersmigration-helpers.md"helper", "shared functions", "loadHelpers"

On-Demand References - Load specific sections:

NeedReferenceWhen
Complete command syntaxcommands.mdQuick Reference insufficient
File format specsmigration-formats.mdCreating files, need exact format
System architecturearchitecture.mdDebugging, explaining internals

Do NOT Load:

  • architecture.md for normal script creation
  • Multiple migration files simultaneously
  • commands.md when Quick Reference suffices

File Locations

TypeLocationExtension
Scripts~/.config/dotrun/scripts/.sh
Aliases~/.config/dotrun/aliases/.aliases
Configs~/.config/dotrun/configs/.config
Helpers~/.config/dotrun/helpers/.sh

Anti-Patterns (NEVER Do)

Script Creation

  • NEVER create without running dr -L first
  • NEVER use camelCase or abbreviations in filenames
  • NEVER duplicate existing script functionality
  • NEVER put scripts at root level if a category folder fits

General

  • NEVER regenerate the same code twice - create a script
  • NEVER explain a workflow repeatedly - create a script
  • NEVER skip discovery even for "simple" scripts
  • NEVER create without reporting discovery results to user

Helper System

Scripts can load reusable modules, they are located in ~/.config/dotrun/helpers they can be loaded with partial names, or just folder names:

[[ -n "${DR_LOAD_HELPERS:-}" ]] && source "$DR_LOAD_HELPERS"
loadHelpers global/colors   # Load by path
loadHelpers my-collection  # Load all from collection

there are all valid imports for the following file: .../helpers/git/pull-requests/getdiffs.sh

loadHelpers git                        # will load all `helpers/**/git/**/*.sh` or `helpers/**/git.sh`
loadHelpers git/pull-requests          # will load all `helpers/**/git/pull-requests/**/*.sh` or helpers/**/git/pull-requests.sh`
loadHelpers pull-requests              # will load all `helpers/**/git/pull-requests/**/*.sh` or helpers/**/git/pull-requests.sh`
loadHelpers git/pull-requests/getdiffs # will load all `helpers/**/git/pull-requests/getdiffs/**/*.sh` or `helpers/**/git/pull-requests/getdiffs.sh`
loadHelpers getdiffs                   # will load all `helpers/**/getdiffs/**/*.sh` or `helpers/**/getdiffs.sh`

Collections for Team Sharing

dr -col add https://github.com/team/scripts.git
dr -col sync   # Check for updates
dr -col update <name>

Pro Tips

  1. Organize: dr set git/cleanup, dr set docker/build
  2. Load order: 01-paths.config before 02-api.config
  3. Document: DOC blocks power dr help and dr -L
  4. Reload: dr -r after config changes
  5. Verify: dr help <name> before running unknown scripts

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

36.32%
按下载量换算35

Claude

30.12%
按下载量换算29

Cursor

19.14%
按下载量换算19

Gemini CLI

8.78%
按下载量换算9

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

可疑

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

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。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。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