Token导航 LogoToken导航

Agent Skill · 研究检索

create-command(创建命令)

create-command 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。

create-command 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为执行命令;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 42,447。

总安装 1,073周安装 43GitHub Stars 7,090下载 347核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

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:create-command(创建命令)
来源仓库:https://github.com/antinomyhq/forge
仓库路径:skills/create-command
安装命令:
npx skills add https://github.com/antinomyhq/forge --skill create-command
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/antinomyhq/forge --skill create-command

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于查找、检索和筛选相关信息,提升内容定位效率。适合在 Codex、Claude、Cursor、Gemini CLI 中根据关键词或任务场景快速获取候选结果。支持创建模块化命令工作流,用于执行特定自动化任务。

安装命令为 npx skills add https://github.com/antinomyhq/forge --skill create-command。所有命令文件必须存放于项目根目录下的 .forge/commands 文件夹内。

来源仓库内路径:skills/create-command

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/antinomyhq/forge --skill create-command;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要执行命令,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

create-command 怎么安装?

在终端执行 npx skills add https://github.com/antinomyhq/forge --skill create-command。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

create-command 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

create-command 需要什么权限?

权限级别是执行命令。安装流程涉及命令执行,可能通过 npx skills add https://github.com/antinomyhq/forge --skill create-command 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

create-command 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

create-command 用的人多吗?

公开数据是累计安装 1,073 次,近一周安装 43 次,下载 347 次,GitHub 7,090 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

create-command 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 来源仓库有 7,090 星,维护活跃度相对可查。

先想清楚的地方

  • 需要执行命令,装到常用环境前要先评估影响面。
  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 不能接受 Skill 在你机器上执行命令的人。
  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
create-command

创建命令 · 本页

1,0737,090执行命令未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Create Commands

Create and manage commands for the code-forge application. Commands are modular workflows that can be invoked to perform specific tasks.

File Location

CRITICAL: All command files must be created in the <cwd>/.forge/commands directory, where <cwd> is the current working directory of your code-forge project.

  • Directory: <cwd>/.forge/commands
  • File format: {command-name}.md
  • Example: If your project is at /home/user/my-project, commands go in /home/user/my-project/.forge/commands/

This is the only location where forge will discover and load custom commands.

Command File Structure

Every command file must have:

  1. YAML Frontmatter (required):

- name: Command identifier (use hyphens for multi-word names) - description: What the command does

  1. Command Body (required):

- List of steps to execute - Special tags for automated workflows - Clear instructions for each step

Example Command File

---
name: check
description: Checks if the code is ready to be committed
---

- Run the `lint` and `test` commands and verify if everything is fine.
  <lint>cargo +nightly fmt --all; cargo +nightly clippy --fix --allow-staged --allow-dirty --workspace</lint>
  <test>cargo insta test --accept --unreferenced=delete</test>
- Fix every issue found in the process

Complete Sample Command

This sample demonstrates all three tag types:

---
name: sample-command
description: Sample command demonstrating the command file structure
---

This is a sample command that demonstrates the structure of command files.

- First step: Perform an initial action
  <lint>echo "Running linting..."</lint>
- Second step: Execute tests
  <test>echo "Running tests..."</test>
- Third step: Complete the workflow
  <shell>echo "Workflow complete!"</shell>
- Final step: Verify everything worked correctly

Creating a New Command

Step 1: Determine Command Purpose

Identify what the command should accomplish:

  • What task will it perform?
  • What steps are involved?
  • Are there automated checks or tests needed?
  • What should the user do with the results?

Step 2: Choose Command Name

Use verb-based names with hyphens for multi-word commands:

  • Good: check, fixme, pr-description, run-tests
  • Bad: checker, fixing, PRdescription

Step 3: Write the Command File

Create the file in the <cwd>/.forge/commands directory with the format: {command-name}.md

IMPORTANT: The file MUST be in <cwd>/.forge/commands where <cwd> is your current working directory. Commands placed anywhere else will not be discovered by forge.

Frontmatter
---
name: your-command-name
description: Clear, concise description of what this command does
---
Command Body

Use markdown lists for steps. Each step should:

  • Start with a clear action verb
  • Be specific and actionable
  • Include context about what to do

Special Command Tags

Use these tags for automated workflows:

<lint> Tag

For linting/formatting commands:

<lint>cargo +nightly fmt --all; cargo +nightly clippy --fix --allow-staged --allow-dirty --workspace</lint>

<test> Tag

For testing commands:

<test>cargo insta test --accept --unreferenced=delete</test>

<shell> Tag

For general shell commands (not linting or testing):

<shell>rm -rf target/debug</shell>

Using Tags

Tags should be placed on their own line after the step description:

- Run linting and testing
  <lint>your-lint-command</lint>
  <test>your-test-command</test>

Command Types

Simple Commands

Single-step or instruction-only commands:

---
name: fixme
description: Looks for all the fixme comments in the code and attempts to fix them
---

Find all the FIXME comments in source-code files and attempt to fix them.

Multi-Step Commands

Commands with multiple sequential steps:

---
name: pr-description
description: Updates the description of the PR
---

- I have created a Pull Request with all the accepted changes
- Understand the current PR deeply using the GH CLI and update the PR title and description
- Make sure the title follows conventional commits standard
- Top-level summary should contain 2-3 lines about the core functionality improvements

Automated Workflow Commands

Commands that include automated checks:

---
name: check
description: Checks if the code is ready to be committed
---

- Run the `lint` and `test` commands and verify if everything is fine.
  <lint>cargo +nightly fmt --all; cargo +nightly clippy --fix --allow-staged --allow-dirty --workspace</lint>
  <test>cargo insta test --accept --unreferenced=delete</test>
- Fix every issue found in the process

Command Templates

Simple Command Template

---
name: simple-command
description: Does one specific thing
---

Single clear instruction or description.

Automated Workflow Template

---
name: automated-workflow
description: Runs automated checks and performs follow-up actions
---

- Run automated checks
  <lint>your-lint-command</lint>
  <test>your-test-command</test>
- Review and fix any issues found
- Complete the workflow

Multi-Step Workflow Template

---
name: multi-step-workflow
description: Performs multiple sequential steps
---

- First step with clear action
- Second step with context
- Third step with specific requirements
- Final step with verification

Git Workflow Template

---
name: git-workflow
description: Performs git operations
---

- Stage changes
  <shell>git add .</shell>
- Run pre-commit checks
  <lint>cargo fmt --all</lint>
  <test>cargo test</test>
- Commit with message
  <shell>git commit -m "your commit message"</shell>
- Push to remote
  <shell>git push</shell>

Best Practices

Naming

  • Use lowercase letters
  • Use hyphens to separate words
  • Use verb-based names (imperative form)
  • Keep names short but descriptive

Descriptions

  • Be clear and concise
  • Describe what the command does, not how
  • Include the main purpose and key outcomes
  • Avoid implementation details

Command Steps

  • Use numbered lists for sequential steps
  • Start each step with an action verb
  • Be specific about what to do
  • Include context for complex steps
  • Use present tense

Special Tags

  • Place tags on their own line after the step
  • Only use <lint>, <test>, and <shell> tags
  • Include complete commands that can be executed
  • Use appropriate flags for your workflow

Common Patterns

Git Workflow Commands

---
name: commit-check
description: Verifies code is ready to commit
---

- Run linting and tests
  <lint>cargo fmt --all; cargo clippy --fix --allow-staged</lint>
  <test>cargo test</test>
- Review and fix any issues
- Stage all changes

Documentation Commands

---
name: update-docs
description: Updates documentation for recent changes
---

- Review recent code changes
- Identify functions or modules that need documentation
- Update inline documentation comments
- Regenerate any auto-generated docs
- Verify documentation builds successfully

Cleanup Commands

---
name: cleanup
description: Cleans up temporary files and artifacts
---

- Remove build artifacts
  <shell>rm -rf target/debug</shell>
- Remove temporary files
  <shell>find . -name "*.tmp" -delete</shell>
- Clean up dependency caches if needed
- Verify the project still builds

Build and Deploy Commands

---
name: build-deploy
description: Builds the project and deploys to staging
---

- Build the project in release mode
  <shell>cargo build --release</shell>
- Run integration tests
  <test>cargo test --test integration</test>
- Build Docker image
  <shell>docker build -t myapp:latest .</shell>
- Tag image for staging
  <shell>docker tag myapp:latest myapp:staging</shell>
- Push to registry
  <shell>docker push myapp:staging</shell>
- Deploy to staging environment
  <shell>kubectl set image deployment/myapp myapp=myapp:staging</shell>

Validation Checklist

Use this checklist to verify your command is complete and correct:

File Structure

  • File is in the <cwd>/.forge/commands directory (CRITICAL)
  • Filename matches command name (e.g., check.md for name: check)
  • File has .md extension
  • YAML frontmatter uses --- delimiters

Frontmatter

  • name field is present
  • name uses lowercase letters
  • name uses hyphens for multi-word names
  • name is verb-based (imperative form)
  • description field is present
  • description is clear and concise
  • description describes what, not how

Command Body

  • At least one step is defined
  • Steps use bullet points (-)
  • Each step starts with action verb
  • Steps are specific and actionable
  • Complex steps include context
  • Steps are in logical order

Special Tags

  • Tags are on their own line after step description
  • Only valid tags are used (<lint>, <test>, <shell>)
  • Tag commands are complete and executable
  • Tag commands use appropriate flags
  • Tag commands are properly formatted

Content Quality

  • Command name is descriptive
  • Steps are clear and unambiguous
  • No redundant or duplicate steps
  • Steps follow logical sequence
  • Special requirements are documented
  • Error handling is considered

Testing

  • Command can be executed successfully
  • All steps complete as expected
  • Special tags work correctly
  • Output is as expected
  • Edge cases are handled
  • Command is recognized by forge: Run forge list command --custom (or forge list cmd) and verify your command appears in the list

Common Mistakes to Avoid

Frontmatter Mistakes

Bad: Wrong delimiter:

---
name: my-command
description: My command

(Missing closing ---)

Good: Correct:

---
name: my-command
description: My command
---

Bad: Missing required field:

---
name: my-command
---

(Missing description)

Good: Correct:

---
name: my-command
description: Does something useful
---

Naming Mistakes

Bad: CamelCase name:

---
name: myCommand
description: Does something
---

Good: Correct:

---
name: my-command
description: Does something
---

Bad: Noun instead of verb:

---
name: checker
description: Checks something
---

Good: Correct:

---
name: check
description: Checks something
---

Step Mistakes

Bad: No action verb:

---
name: test
description: Runs tests
---

- The tests
- The code

Good: Correct:

---
name: test
description: Runs tests
---

- Run all tests
- Verify code quality

Bad: Vague steps:

---
name: deploy
description: Deploys application
---

- Do the deployment
- Make sure it works

Good: Correct:

---
name: deploy
description: Deploys application to production
---

- Build the Docker image
  <shell>docker build -t myapp:latest .</shell>
- Push to registry
  <shell>docker push myapp:latest</shell>
- Deploy to production
  <shell>kubectl set image deployment/myapp myapp=myapp:latest</shell>
- Verify deployment is healthy

Tag Mistakes

Bad: Tag on same line:

- Run tests <test>cargo test</test>

Good: Correct:

- Run tests
  <test>cargo test</test>

Bad: Invalid tag:

- Run checks
  <check>cargo clippy</check>

Good: Correct:

- Run checks
  <lint>cargo clippy</lint>

Bad: Incomplete command:

- Format code
  <lint>cargo fmt</lint>

(Missing --all flag)

Good: Correct:

- Format code
  <lint>cargo fmt --all</lint>

Quick Reference

File Location

  • Directory: <cwd>/.forge/commands (where <cwd> is current working directory)
  • Format: {command-name}.md
  • CRITICAL: Commands MUST be in this exact location to be discovered by forge

Valid Tags

  • <lint> - For linting/formatting commands
  • <test> - For testing commands
  • <shell> - For general shell commands

Naming Rules

  • Lowercase only
  • Hyphens for multi-word names
  • Verb-based (imperative form)
  • Keep it short but descriptive

Step Guidelines

  • Start with action verb
  • Be specific
  • Include context for complex steps
  • Use present tense
  • Keep steps focused

When to Use Tags

  • Use <lint> when running formatters or linters
  • Use <test> when running test suites
  • Use <shell> for other shell commands
  • Place tags on their own line after step description
  • Don't use tags if the step is just an instruction

Testing Your Command

After creating a command, test it by:

  1. Syntax Check: Verify YAML is valid # If you have yamllint installed yamllint path/to/your-command.md
  2. Manual Review: Read through the command

- Does each step make sense? - Is the order logical? - Are all commands complete?

  1. Execution Test: Run the command

- Does each step execute successfully? - Is the output as expected? - Are there any errors?

  1. Forge Recognition Test: Verify the command is recognized by forge # Option 1: List all commands (custom commands marked as type: custom) forge list command # Option 2: List only custom commands forge list cmd # Option 3: List only custom commands (newer versions) forge list command --custom

- Does your command appear in the list? - Is the name correct? - Is the description correct?

  1. Edge Cases: Consider unusual scenarios

- What happens if a step fails? - What if the environment is different? - What if files are missing?

Verification

After creating a command:

  1. Verify the file location: Ensure the file is in <cwd>/.forge/commands directory (CRITICAL - commands anywhere else will not be found)
  2. Check YAML frontmatter is valid (use --- delimiters)
  3. Ensure the command name matches the filename (without.md)
  4. Verify the command is recognized by forge: # Option 1: List all commands (custom commands marked as type: custom) forge list command # Option 2: List only custom commands forge list cmd # Option 3: List only custom commands (newer versions) forge list command --custom Your new command should appear in the list with its name and description
  5. Test the command to ensure it works as expected
  6. Verify special tags are properly formatted

If your command doesn't appear in the list, check:

  • File location: File MUST be in <cwd>/.forge/commands directory (this is the most common issue)
  • Filename matches the name field in frontmatter
  • YAML frontmatter is properly formatted with --- delimiters
  • Both name and description fields are present

Getting Help

If you're unsure about something:

  • Review the examples in this skill
  • Follow the validation checklist
  • Test your command before finalizing

平台分布

Codex

36.37%
按下载量换算126

Claude

28.9%
按下载量换算100

Cursor

21.17%
按下载量换算73

Gemini CLI

8.89%
按下载量换算31

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

执行命令

安装流程涉及命令执行,可能通过 npx skills add https://github.com/antinomyhq/forge --skill create-command 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

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。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills