new

Creates OpenClaw components with correct directory placement, naming conventions, and code templates.

Updated Mar 30, 2026
One-click install
npx skills add https://github.com/guizonatto/myopen-claw --skill new-guizonatto
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: new
Source: https://github.com/guizonatto/myopen-claw/tree/main/.claude/skills/new
Command: npx skills add https://github.com/guizonatto/myopen-claw --skill new-guizonatto

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When adding new code to an OpenClaw project, developers must remember where each component type belongs, which naming convention applies, and what boilerplate structure is required. This Skill removes that guesswork by scaffolding skills, tools, pipes, crons, triggers, and agents in the right location with the right template. ## Core Features & Use Cases - Component Type Inference: Detects the intended component type from natural language descriptions (e.g., "run every day" maps to cron, "send a message" maps to tool). - Location & Naming Rules: Places files in the correct directory (skills/, tools/, pipes/, crons/, triggers/, agents/) following the {dominio}_{acao}.py naming convention, including subdirectory rules for domains with 3+ skills. - Mandatory Templates: Generates boilerplate with header docstrings (Função, Usar quando, ENV_VARS, DB_TABLES), standalone __main__ execution blocks, and no horizontal imports between same-layer components. - Use Case: Ask to create a component that fetches news daily and posts it to Telegram, and the Skill scaffolds the skill, pipe, and cron files with correct templates, then reminds you to update .env.example and add tests. ## Quick Start Ask the assistant to create a new OpenClaw component, for example: create a cron that runs the lead capture pipe every morning at 8am.

Frequently Asked Questions about new

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create a new OpenClaw skill or tool file?

Describe the behavior you want in natural language and the Skill infers the component type, then creates the file in the correct directory with the proper naming convention and template. For example, "send a notification" creates a tool, while "fetch data from X" creates a skill.

How does the Skill decide between skill, tool, pipe, cron, trigger, and agent?

It maps intent keywords to types: fetching data maps to skill, sending messages to tool, scheduled recurring runs to cron, webhook or command events to trigger, conditional decision logic to agent, and connecting a skill to a tool maps to pipe.

What naming convention do OpenClaw component files follow?

Files follow the pattern `{dominio}_{acao}.py` for skills, `{canal}_notify.py` or `{canal}_scraper.py` for tools, `{processo}_pipe.py` for pipes, `{processo}_cron.py` for crons, `{canal}_trigger.py` for triggers, and `{nome}_agent.py` for agents.

When should skills be moved into a subdirectory?

Keep skills flat in the `skills/` directory when a domain has up to two skills. Once a domain reaches three or more skills, create a subdirectory `skills/{dominio}/` containing an `__init__.py` file.

What boilerplate must every OpenClaw component include?

Every component requires a header docstring with Função, Usar quando, ENV_VARS, and DB_TABLES sections, plus an `if __name__ == "__main__"` block for standalone execution. Horizontal imports between same-layer components are prohibited.

What happens after the component file is created?

The Skill checks whether new environment variables must be added to `.env.example`, creates a test file under `tests/{layer}/`, reports the exact file path, and offers to scaffold related components for a complete process.