When should I use semantic query construction?

Use it when you need to capture the conceptual meaning of a request, such as 'user preferences for error handling,' rather than using generic or temporal keywords like 'error handling' or 'what we did yesterday.'

What are the different memory types supported?

The skill categorizes memories into 'episodic' for raw observations, 'semantic' for synthesized insights, and 'procedural' for specific workflows and how-to knowledge.

How does the weight tuning work in Engram?

It uses a three-factor scoring system based on Recency, Importance, and Relevance, which can be dynamically adjusted depending on whether the intent is to find recent context or long-term best practices.

What is the purpose of the Memory Retrieval skill?

It provides specialized patterns for searching an AI's memory graph to ensure Claude retrieves the most relevant and useful historical context for its current tasks.

How does the skill handle insufficient search results?

It provides strategies to broaden the query semantically, reduce relevance weights to boost other factors, or remove specific type filters to find the next best matches.

Engram Memory Retrieval

Name: Engram Memory Retrieval
Author: grootenberg

bygrootenberg

0•

生産性とワークフロー

Optimizes how Claude searches and synthesizes information from memory graphs to provide contextually accurate responses.

The Memory Retrieval skill for the Engram plugin empowers Claude to navigate complex memory graphs using advanced query formulation and weighted scoring strategies. By leveraging semantic search patterns, three-factor weight tuning—balancing recency, importance, and relevance—and specific memory type filtering, it ensures that the most pertinent historical context, decisions, and patterns are surfaced during development. This skill is essential for maintaining long-term project continuity, recalling specific architectural decisions, and ensuring that Claude's responses are grounded in established project knowledge rather than generic assumptions.

主な機能

01Memory Type Filtering for episodic, semantic, and procedural data

02Query Decomposition strategies for complex multi-part retrievals

03Semantic Query Construction for conceptual meaning over keywords

04Context Assembly patterns with result synthesis and scoring thresholds

05Three-Factor Scoring based on recency, importance, and relevance

060 GitHub stars

ユースケース

01Synthesizing best practices from historical project interactions

02Recalling project-specific architectural decisions and coding patterns

03Retrieving specific procedural workflows for complex development tasks

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add grootenberg/engram memory-retrieval

For use in Claude.ai and ChatGPT

Download Skill