Does it support parallel processing?

Yes, it can spawn up to 5 parallel agents (using the Haiku model) to groom multiple items simultaneously, significantly speeding up the refinement process.

Can I groom specific sections of my backlog?

Yes, you can pass section names like 'P1', 'P2', or 'Ideas' as arguments to groom all items within that specific priority group.

Where are the grooming results stored?

The skill provides a summary in the chat and offers to save detailed reports in the .claude/grooming-reports/ directory for future reference.

How does this skill find my backlog items?

The skill specifically looks for a file located at .claude/BACKLOG.md in your repository and parses it for item titles and priority sections.

What kind of output does the groomer provide?

It generates a comprehensive Markdown report including a summary table, individual context manifests for each item, and analysis of cross-item dependencies.

Backlog Item Groomer

Name: Backlog Item Groomer
Author: Jamie-BitFlight

byJamie-BitFlight

•

Herramientas de Colaboración

Prepares project backlog items for implementation by discovering relevant research, required skills, and technical context using parallel agents.

The Backlog Item Groomer skill automates the refinement of your project tasks by parsing the .claude/BACKLOG.md file and gathering essential context for development. It intelligently identifies target items by title, priority section, or bulk selection, then spawns parallel sub-agents to research prior work and dependencies. This process transforms vague requirements into actionable developer manifests, complete with success criteria, shared dependencies, and suggested groupings, ensuring that every task is fully prepared before coding begins.

Características Principales

0118 GitHub stars

02Automated parsing of .claude/BACKLOG.md files and priority sections

03Automated identification of shared dependencies and research gaps

04Detailed grooming report generation with cross-item findings

05Comprehensive context discovery for research, skills, and plugins

06Parallel agent execution for rapid multi-item grooming

Casos de Uso

01Identifying technical blockers and shared dependencies across a large set of backlog items

02Bulk grooming an entire priority section (e.g., P1) to prepare for a new development cycle

03Preparing a specific high-priority task for immediate development by gathering all technical context

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jamie-bitflight/claude_skills groom-backlog-item

For use in Claude.ai and ChatGPT

Download Skill