qwen_cleanup_strategist

Group Gemma noise-labeled files into batches and generate a cleanup plan with MPS scores.

2|Updated Mar 28, 2025
One-click install
npx skills add https://github.com/Foundup/Foundups-Agent --skill qwen-cleanup-strategist
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: qwen_cleanup_strategist
Source: https://github.com/Foundup/Foundups-Agent/tree/main/.claude/skills/qwen_cleanup_strategist_prototype
Command: npx skills add https://github.com/Foundup/Foundups-Agent --skill qwen-cleanup-strategist

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill enables automated planning of safe, structured cleanup operations by analyzing Gemma's noise labels and applying WSP safety rules to ensure no critical data is touched.

Core Features & Use Cases

  • Group high-confidence noise files by category to enable targeted cleanup.
  • Apply WSP safety checks (83/64) and flag any violations for 0102 review.
  • Produce a structured cleanup plan with batch metadata, rationale, and execution readiness.

Quick Start

Run the strategist on data/gemma_noise_labels.jsonl to generate a cleanup plan.

Frequently Asked Questions about qwen_cleanup_strategist

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

FAQPage Schema
How do I plan a safe cleanup batch for noise-labeled files?

To plan a safe cleanup batch, you can group Gemma's noise-labeled files by category and apply WSP 83/64 safety checks to generate a structured plan with batching up to 50 files.

What are WSP 83/64 safety checks in data cleanup?

WSP 83/64 safety checks are validation rules applied during cleanup planning to ensure no critical data is touched, automatically flagging any rule violations for 0102 review.

How do I triage Gemma noise labels for automated file removal?

You can triage Gemma noise labels by analyzing high-confidence files and grouping them by category, which enables targeted cleanup while applying WSP safety rules to protect data.

Can I process more than 50 noise files in a single cleanup batch?

Batching is limited to up to 50 files per batch to ensure safe, structured cleanup operations, meaning larger sets of noise-labeled files are divided into multiple structured batches.

What is an MPS score in a cleanup plan?

An MPS score is a metric included in the generated cleanup plan that helps evaluate and prioritize the noise-labeled files selected for safe removal.

How do I generate a cleanup plan from a gemma_noise_labels.jsonl file?

You generate a cleanup plan by running the strategist on the data/gemma_noise_labels.jsonl file, which outputs a structured data/cleanup_plan.json file along with an execution_id.