automate

Convert task descriptions into headless, re-runnable coding-agent CLI scripts.

7|2|Updated Mar 5, 2026
One-click install
npx skills add https://github.com/AndurilCode/craftwork --skill automate-andurilcode
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: automate
Source: https://github.com/AndurilCode/craftwork/tree/main/plugins/craftwork-professional/skills/automate
Command: npx skills add https://github.com/AndurilCode/craftwork --skill automate-andurilcode

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps you turn coding-agent prompts into reliable, repeatable automation that can run headlessly in CI/cron or dispatch cloud agent tasks to open pull requests asynchronously.

Core Features & Use Cases

  • Build deterministic automation around agent CLIs: orchestrate Claude Code, Codex, Gemini CLI, Copilot CLI, or GitHub gh agent-task in non-interactive mode with structured outputs and re-runnable contracts.
  • Select a safe execution model: choose local execution for synchronous pipelines or cloud delegation when you want a PR produced without blocking.
  • Control safety and observability: enforce non-negotiable flags (auto mode, tool scope allowlists, JSON output/schema, success verification) and emit session IDs and cost/diagnostics for downstream tooling.

Quick Start

Tell the automation engine what you want to run (for example: automate a review-then-fix workflow) and specify that it should execute using a coding-agent CLI in headless mode and output structured JSON suitable for piping into jq.

Frequently Asked Questions about automate

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

FAQPage Schema
How do I automate coding-agent CLI workflows in a CI pipeline?▼

Automate coding-agent CLI workflows in a CI pipeline by converting task descriptions into headless, re-runnable scripts. This enforces deterministic auto-mode flags, tool allowlists, and structured JSON outputs, enabling non-interactive execution for automated code generation and review loops.

Can I run Claude Code or Gemini CLI headlessly for automated code refactoring?▼

Yes, you can run Claude Code, Gemini CLI, Codex, or Copilot CLI headlessly for automated code refactoring. The automation engine orchestrates these agents in non-interactive mode with structured outputs and re-runnable contracts to ensure reliable execution.

What is the best way to convert a ticket into a GitHub PR using a coding agent?▼

The best way to convert a ticket into a GitHub PR is using a ticket-to-PR pipeline that dispatches cloud agent tasks asynchronously. This approach delegates the code generation to a cloud agent, producing a pull request without blocking your local execution environment.

Does headless agent execution support tool allowlists and structured JSON output?▼

Headless agent execution supports tool allowlists and structured JSON output by enforcing non-negotiable flags for auto mode and tool scope. This ensures safe execution boundaries and provides schema-based outputs suitable for piping into tools like jq.

How do I verify the success of an automated coding-agent task in a cron job?▼

Verify the success of an automated coding-agent task in a cron job using JSON-based success verification. The engine emits session IDs and cost tracking diagnostics, allowing downstream tooling to confirm execution status and observability for your scheduled tasks.