How does the skill handle search failures?

It uses predefined fallback chains to automatically try alternative search methods if the primary layer returns no results, times out, or fails a health check.

Does this skill perform the actual data search?

No, it is a policy and routing engine. It works alongside skills like memory-query, vector-search, and bm25-search to execute the actual retrieval operations it recommends.

What is an explainability payload?

It is a detailed JSON report returned after every retrieval that lists which layers were tried, which ones succeeded, the execution time, and the confidence level of the results.

What is the primary purpose of the Retrieval Policy skill?

It acts as a decision-making 'brainstem' that analyzes a search query's intent and decides which specific search methods (like vector, keyword, or topic-based) will yield the best results.

Intelligent Memory Retrieval Policy

Name: Intelligent Memory Retrieval Policy
Author: SpillwaveSolutions

bySpillwaveSolutions

•

데이터 과학 및 ML

Optimizes agent memory searches by automatically classifying query intent and routing requests through the most efficient retrieval layers.

The Retrieval Policy skill serves as the cognitive decision-making engine for agent memory systems, determining the optimal strategy for retrieving information based on real-time query analysis. By managing a multi-tiered architecture that spans topic graphs, hybrid search, vector embeddings, and BM25 keyword matching, it ensures agents access the most relevant data with maximum efficiency. It features robust fallback chains to handle layer failures and provides a detailed explainability payload, allowing developers to understand exactly how and why specific information was retrieved.

주요 기능

01Multi-layer fallback chains to ensure search reliability during service outages

02Intent-specific search constraints including max depth, beam width, and timeouts

03Automated query intent classification for Explore, Answer, Locate, and Time-boxed intents

04Dynamic routing across five capability tiers from Agentic to Full Hybrid search

05Comprehensive explainability payloads reporting tiers used, layers tried, and confidence scores

064 GitHub stars

사용 사례

01Debugging agent reasoning by inspecting the retrieval logic and layer-by-layer search performance

02Optimizing search precision and performance for large-scale agent knowledge bases

03Implementing graceful degradation for AI agents when vector databases or specific search layers are unavailable

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add spillwavesolutions/agent-memory retrieval-policy

For use in Claude.ai and ChatGPT

Download Skill