create-dedup-task

Generate a deduplication task that scans for exact and near-duplicate papers and suggestions.

16|8|Updated Apr 15, 2026
One-click install
npx skills add https://github.com/GliteTech/glite-arf --skill create-dedup-task
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: create-dedup-task
Source: https://github.com/GliteTech/glite-arf/tree/main/arf/skills/create-dedup-task
Command: npx skills add https://github.com/GliteTech/glite-arf --skill create-dedup-task

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The Dedup Task Creator provides a structured approach to generate a new deduplication workflow that scans for exact duplicates across tasks and overlapping work, enabling automated cleanup and corrections within a multi-task project.

Core Features & Use Cases

  • Scan tasks//assets/paper/ to detect papers with identical paper_id across tasks.
  • Run cross-paper similarity checks (DOI or near-duplicate titles) to surface duplicates.
  • Compare active suggestions across tasks to identify overlaps and remediation opportunities.
  • Determine the canonical copy (earliest date_added, then lowest task_id) and generate paper corrections to delete extras.
  • Create suggestion corrections to reject duplicates and ensure consistency across the project.
  • Validate corrections with verify_corrections and re-run aggregators to confirm changes are applied.
  • Use this skill when cleaning up duplicated work or overlaps in a multi-task research project.

Quick Start

Follow the /create-task skill instructions to create the new dedup task folder using the provided description.

Frequently Asked Questions about create-dedup-task

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

FAQPage Schema
How do I automate deduplication across multiple research tasks?▼

Automating deduplication across research tasks requires generating a structured task that scans for exact duplicates and overlapping work, then prepares corrections to ensure consistent artifact hygiene. This removes manual checks in multi-task projects.

How does cross-task similarity checking work for research papers?▼

Cross-task similarity checking works by scanning aggregated task assets to detect identical paper IDs, matching DOIs, or near-duplicate titles. It surfaces overlapping work so you can remediate duplicated research artifacts effectively.

How do I determine the canonical copy when resolving duplicate papers?▼

To determine the canonical copy when resolving duplicate papers, the system selects the record with the earliest date added, using the lowest task ID as a tiebreaker. It then generates corrections to delete the extra copies.

How do I validate corrections and re-run aggregators after a deduplication task?▼

To validate corrections after a deduplication task, you run verify_corrections on the generated outputs and then re-run aggregators. This confirms that the paper and suggestion corrections are properly applied across tasks.

Can I use this for cleaning up overlapping suggestions in a multi-task project?▼

Yes, you can use this to clean up overlapping suggestions in a multi-task project. It compares active suggestions across tasks to identify overlaps and creates corrections to reject duplicates, ensuring project-wide consistency.