Execute qualquer Skill no Manus
com um clique

Execute qualquer Skill no Manus com um clique

create-command

Estrelas33

Forks3

Atualizado12 de junho de 2026 às 14:06

Create new direct commands, modify existing commands. Use when users want to create a deterministic command that is directly executed by the editor via a shell script.

Instalação

Instalar com Codex ou Claude Copie este prompt, cole no Codex, Claude ou outro assistente e deixe que ele revise a página da skill e instale para você.

Executar no Manus

Fonte

w-t-yang

w-t-yang/aynite

Abrir repositório GitHub Ver repositórios do creator

Download

Executar no Manus

Ocupações relacionadasSOC

Baseado na classificação ocupacional SOC

Desenvolvedores de softwareInformática e Matemática·SOC 15-1252

SKILL.md

readonly

name	create-command
description	Create new direct commands, modify existing commands. Use when users want to create a deterministic command that is directly executed by the editor via a shell script.

Command Creator

A skill for creating new "Direct Commands" and iteratively improving them.

Output Location

Before creating any files, confirm the output folder with the user. If the user doesn't specify one:

All platforms: Commands are stored in ~/.aynite/commands/

Each command lives in its own subdirectory: <output-folder>/<command-name>/

Command Structure

A direct command MUST follow this exact structure:

command-name/
├── COMMAND.md     - Documentation with YAML parameters
├── run.sh         - THE ENTRY POINT (required)
└── scripts/       - Support scripts (Python, JS, etc.)

Anatomy of COMMAND.md

The COMMAND.md file must start with a YAML frontmatter block defining its parameters:

---
name: my-command
description: What this command does
parameters:
  - name: input_file
    description: The file to process
    required: true
  - name: mode
    description: Processing mode (fast|slow)
    default: fast
---

# My Command Documentation
Detailed instructions on how to use this command...

The run.sh Entry Point

The run.sh script is the primary execution point. It should:

Parse parameters passed by the editor.
Delegate to scripts in the scripts/ folder if necessary (e.g., python3 scripts/process.py).
Provide clear output or error messages.

Creation Workflow

Confirm the output folder — Ask the user where to create the command. If they don't have a preference, use the default Aynite commands folder (~/.aynite/commands/ on all platforms). Create the folder if it doesn't exist.
Define Interface — What parameters does this command need?
Implement Logic — Write the run.sh and any supporting scripts in scripts/. Make sure run.sh is executable (chmod +x run.sh).
Document — Fill out the COMMAND.md with the parameter schema and descriptions.
Tell the user what you created — After creation, clearly tell the user:
- The name of the command and where it was saved
- How to invoke it: type > <command-name> in the chat (e.g., > my-command)
- Important: They should press Ctrl+R to refresh the tile so Aynite discovers the new command. Until they reload, the command won't appear in the command palette or be invocable.

Mais deste repositório

mesmo repositório

transform-to-dataview

w-t-yang/aynite

Transform raw data (files, folders, clipboard content) into JSON files compatible with Aynite's dataview visualization types. When the user mentions wanting to visualize data, create a chart/graph/diagram/flow/mindmap from data, convert data to a specific format, prepare data for a view, or transform spreadsheets, CSVs, or structured text into a visual format — use this skill. Also use it when the user has data they want to "render in a dataview" or "make a dataview from this." This skill helps brainstorm the best visualization approach, handles the transformation, validates the output against the view's schema, and saves the result.

2026-06-2333

create-skill

w-t-yang/aynite

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.

2026-06-1233

create-spell

w-t-yang/aynite

Create a new "spell" in Aynite. A spell is a unified concept — it can be either a command (deterministic script, invoked with >) or a skill (AI-guided workflow, invoked with /). Use this skill whenever a user wants to create a spell, cast magic, automate something, or build a new capability but isn't sure whether it should be a command or a skill. Also use when users say "I want to make a spell", "teach Aynite to...", "I wish Aynite could...", or describe a task they want to automate or delegate.

2026-06-1233

create-theme

w-t-yang/aynite

Create custom Aynite themes. Use this skill when a user wants to create a new color theme, customize their editor appearance, design a dark/light theme, or generate a theme based on a mood, image, or color palette. Also use when users mention "theme", "color scheme", "appearance", or "look and feel" of the editor.

2026-06-1233

create-slides

w-t-yang/aynite

Create interactive slide decks using reveal.js. Use when users want to create presentations, slide decks, slide shows, or any kind of presentation materials. Triggers on phrases like "create slides", "make a presentation", "turn this into slides", "build a slide deck", or any request involving presenting content in slide format. Also triggers when users provide structured content (markdown, notes, outlines) and want it rendered as visual slides. If the user has content files (markdown, text), suggest using this skill to turn them into an interactive reveal.js presentation.

2026-05-2733

Command Creator

A skill for creating new "Direct Commands" and iteratively improving them.

Output Location

Before creating any files, confirm the output folder with the user. If the user doesn't specify one:

All platforms: Commands are stored in ~/.aynite/commands/

Each command lives in its own subdirectory: <output-folder>/<command-name>/

Command Structure

A direct command MUST follow this exact structure:

command-name/ ├── COMMAND.md - Documentation with YAML parameters ├── run.sh - THE ENTRY POINT (required) └── scripts/ - Support scripts (Python, JS, etc.)

Anatomy of COMMAND.md

The COMMAND.md file must start with a YAML frontmatter block defining its parameters:

--- name: my-command description: What this command does parameters: - name: input_file description: The file to process required: true - name: mode description: Processing mode (fast|slow) default: fast --- # My Command Documentation Detailed instructions on how to use this command...

The run.sh Entry Point

The run.sh script is the primary execution point. It should:

Parse parameters passed by the editor.

Delegate to scripts in the scripts/ folder if necessary (e.g., python3 scripts/process.py).

Provide clear output or error messages.

Creation Workflow

Confirm the output folder — Ask the user where to create the command. If they don't have a preference, use the default Aynite commands folder (~/.aynite/commands/ on all platforms). Create the folder if it doesn't exist.

Define Interface — What parameters does this command need?

Implement Logic — Write the run.sh and any supporting scripts in scripts/. Make sure run.sh is executable (chmod +x run.sh).

Document — Fill out the COMMAND.md with the parameter schema and descriptions.

Tell the user what you created — After creation, clearly tell the user:

The name of the command and where it was saved
How to invoke it: type > <command-name> in the chat (e.g., > my-command)
Important: They should press Ctrl+R to refresh the tile so Aynite discovers the new command. Until they reload, the command won't appear in the command palette or be invocable.