How does it detect duplicates in the database?

The skill uses a tight similarity threshold (0.25) to compare entries, ensuring that only highly similar or identical files are flagged as duplicate clusters.

Does this skill automatically delete files?

No, this is a discovery tool. It identifies candidates for consolidation. To fetch content or perform actions on these files, you should use companion skills like consolidate-actions.

Why is the output format so compact?

The output is designed as compact JSON to prevent 'token bloat,' ensuring that the AI can review the discovery results without consuming significant parts of its context window.

Can I limit the number of items returned?

Yes, you can use the --limit parameter to set the maximum number of items per category (duplicates or outdated), which defaults to 20 to keep output concise.

What counts as 'outdated' content?

The tool scans for specific markers, such as 'temporary', within the file metadata or content to identify notes or fixes that were intended to be transient.

Consolidate Discovery

Name: Consolidate Discovery
Author: Connsulting

byConnsulting

•

Database Management

Identifies duplicate and outdated content within SQLite databases to streamline knowledge bases and reduce token usage.

Consolidate Discovery is a specialized utility designed to audit and maintain knowledge repositories stored in SQLite databases. By analyzing file similarity with a tight threshold and scanning for metadata markers like 'temporary', it flags redundant entries and legacy content that may clutter your workflow. This skill is specifically optimized for AI-driven development by providing a compact JSON output that minimizes token consumption, allowing Claude to efficiently identify candidates for consolidation or removal without exhausting the context window.

Key Features

011 GitHub stars

02Identification of outdated content via metadata markers

03Automated duplicate detection using tight similarity thresholds

04Token-optimized compact JSON output for efficient AI processing

05Direct integration with SQLite-based learning and documentation repositories

06Configurable discovery modes for targeted database auditing

Use Cases

01Cleaning up redundant documentation within a project's knowledge base

02Pruning temporary notes and outdated markers from a learning database

03Preparing database content for AI context ingestion by reducing data bloat

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add connsulting/claude-code-plugins consolidate-discovery

For use in Claude.ai and ChatGPT

Download Skill