shovel-ready

Drive GitHub shovel-ready labeled issues through TDD to merged pull requests.

Updated Feb 23, 2026
One-click install
npx skills add https://github.com/eumemic/dev-skills --skill shovel-ready
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: shovel-ready
Source: https://github.com/eumemic/dev-skills/tree/main/skills/shovel-ready
Command: npx skills add https://github.com/eumemic/dev-skills --skill shovel-ready

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill turns a human-maintained GitHub shovel-ready issue label queue into a reliable, end-to-end autonomous flow that drives issues through TDD, /ship, /retro, and merge completion while optionally auditing and refilling the queue when it’s empty.

Core Features & Use Cases

  • Label-gated queue execution: Processes only issues that have the shovel-ready label, preserving a deliberate readiness gate and preventing speculative work.
  • TDD-first implementation per issue: For each selected issue, writes failing tests first, confirms the expected failure, implements the minimal fix/feature, and ships via the shared pipeline.
  • Empty-queue audit + refill: When no labeled issues exist, audits open work to find closure candidates and unlabeled near-ready issues, then asks for confirmation before labeling or closing.
  • Wait-mode long-cadence monitoring: Re-enters on a schedule when the queue remains empty, with idle streak escalation and optional autonomous merge.

Quick Start

Ask your AI agent to run: /shovel-ready and automatically work the shovel-ready issue queue until the next merge-ready outcome is reached.

Frequently Asked Questions about shovel-ready

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

FAQPage Schema
How do I automate GitHub issue triage to merged pull requests?

You can automate this flow by maintaining a `shovel-ready` label queue on GitHub issues. The system detects labeled issues via GitHub CLI, implements fixes using a TDD-first workflow, and optionally merges pull requests autonomously after CI passes.

What is a TDD-first autonomous development loop for GitHub issues?

A TDD-first autonomous development loop processes queued GitHub issues by writing failing tests first, confirming the expected failure, implementing the minimal fix, and shipping via a pipeline. Queue state routing determines whether to execute, audit, or wait.

How do I manage an empty queue in autonomous development workflows?

When an autonomous development queue is empty, the system audits open GitHub issues to find closure candidates and unlabeled near-ready issues. It requests confirmation before labeling or closing, re-entering on a schedule with idle streak escalation.

Does autonomous pull request merging work with GitHub CLI and CI monitoring?

Autonomous pull request merging works with GitHub CLI by detecting queue states and monitoring CI. After CI passes on a pull request created through the TDD workflow, the system can optionally merge it automatically using safe PR closing semantics.

Can I use label-gated execution to prevent speculative work on GitHub issues?

Label-gated execution processes only GitHub issues with a specific readiness label, preserving a deliberate gate that prevents speculative work. This ensures the autonomous workflow acts solely on issues explicitly marked as ready for development.

What are the limitations of autonomous issue triage when the queue is empty?

When the autonomous issue triage queue is empty, the system cannot execute new work without human confirmation. It enters a wait-mode for long-cadence monitoring, escalating idle streaks and requiring manual approval before labeling or closing unlabeled near-ready issues.