curate

Score uncurated entities, merge aliases, and prune noise in a custom knowledge graph.

1|Updated Dec 28, 2025
One-click install
npx skills add https://github.com/JDHayesBC/Awareness --skill curate-jdhayesbc
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: curate
Source: https://github.com/JDHayesBC/Awareness/tree/main/.claude/skills/curate
Command: npx skills add https://github.com/JDHayesBC/Awareness --skill curate-jdhayesbc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Curate passes address uncurated entities and edges in the custom knowledge graph, scoring importance, merging aliases, pruning noise, and enforcing TTL for technical artifacts to maintain coherent identity across sessions.

Core Features & Use Cases

  • Score uncurated entities and edges by importance
  • Merge aliases and resolve aliases into canonical entities
  • Prune low-signal or noisy relationships while preserving critical links
  • Enforce tech TTL and lifecycle rules to keep the graph current

Quick Start

Run the curate pass during reflection cycles to clean and normalize the knowledge graph.

Frequently Asked Questions about curate

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

FAQPage Schema
How do I clean and enrich a knowledge graph with uncurated entities?

Knowledge graph curation cleans and enriches data by scoring uncurated entities, merging aliases, and pruning noise. This process maintains coherent identity and relationships across multiple sessions and contexts.

What is the best way to resolve aliases into canonical entities in a Cypher graph?

Resolving aliases into canonical entities in a Cypher graph involves merging duplicate entries during reflection cycles. This normalization ensures persistent AI memory maintains a coherent identity across sessions.

How does TTL enforcement work for technical artifacts in a knowledge graph?

TTL enforcement for technical artifacts in a knowledge graph prunes expired data and applies lifecycle rules. This keeps the graph current by automatically removing outdated technical artifacts during curation passes.

Can I safely test knowledge graph updates without risking data loss?

You can safely test knowledge graph updates using dry-run mode and mandatory group_id filtering. These guardrails ensure reproducible, auditable changes by enforcing safe writes before modifying the graph.

When should I run a curation pass on my knowledge graph?

You should run a curation pass during reflection cycles to clean and normalize the knowledge graph. This timing ensures uncurated entities, noisy relationships, and technical artifacts are processed systematically.