decay-scoring

Compute obsolescence scores for RAG corpus knowledge nodes using configurable weights.

Updated Jan 15, 2026
One-click install
npx skills add https://github.com/arbgjr/smart_alarm --skill decay-scoring
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: decay-scoring
Source: https://github.com/arbgjr/smart_alarm/tree/main/.claude/skills/decay-scoring
Command: npx skills add https://github.com/arbgjr/smart_alarm --skill decay-scoring

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires yaml, and includes scripts (resource) components.

What problem does it solve?

The decay-scoring skill computes a quantitative obsolescence score for knowledge nodes in a RAG corpus. It helps teams identify outdated or under-validated content to prioritize review, update, or archiving.

Core Features & Use Cases

  • Automatic age, validation recency, and access-based decay scoring to surface stale content.
  • Configurable weights, half-lives, and content-type bonuses to tailor prioritization.
  • Generates per-node scores, status classifications, and a review queue for curator actions.
  • Use Case: a growing knowledge base flags items for periodic review, validation, or archiving to maintain freshness.

Quick Start

Run the decay scoring workflow using the included scripts:

  • python3 .claude/skills/decay-scoring/scripts/decay_calculator.py
  • python3 .claude/skills/decay-scoring/scripts/decay_tracker.py access NODE_ID
  • python3 .claude/skills/decay-scoring/scripts/decay_trigger.py

Frequently Asked Questions about decay-scoring

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

FAQPage Schema
How do I identify outdated content in a RAG knowledge base for review or archiving?

Decay scoring identifies outdated content in a RAG knowledge base by computing quantitative obsolescence scores across age, validation recency, access frequency, and content type to surface stale nodes for curator actions.

What is the best way to automate knowledge curation metrics for stale corpus nodes?

Automating knowledge curation metrics involves applying configurable weights and half-lives to corpus nodes, generating per-node obsolescence scores, status classifications, and a review queue to prioritize updates or archiving.

Can I customize the decay scoring weights for different content types in my corpus?

Yes, you can customize decay scoring by adjusting configurable weights, half-lives, and content-type bonuses defined in YAML metadata to tailor content prioritization across your specific knowledge base nodes.

How do I track access frequency for RAG nodes to trigger decay calculations?

You track access frequency for RAG nodes by running the decay_tracker script with node access events, feeding node history and decay parameters into the decay calculator to trigger automated scoring updates.

Does this automated decay scoring approach work with large scale knowledge bases?

Yes, the automated decay scoring approach works with large scale knowledge bases, applying configurable decay parameters across age, validation recency, access frequency, and content type to prioritize items for review at scale.