skills.rest
ehfazrezwan
ehfazrezwanCOMMUNITY
Explore all skills by this author →
GitHub

search

Retrieve relevant stored memories using hybrid vector and temporal knowledge graph search.

5|1|Updated Feb 10, 2026
Search Tags:#knowledge graph#MCP#project context#hybrid search#memory search#neuralscape#recall memories
One-click install
npx skills add https://github.com/ehfazrezwan/neuralscape --skill search-ehfazrezwan
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: search
Source: https://github.com/ehfazrezwan/neuralscape/tree/main/neuralscape-plugin/skills/search
Command: npx skills add https://github.com/ehfazrezwan/neuralscape --skill search-ehfazrezwan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Users lose critical context like past preferences, project decisions, and discussed facts between AI assistant sessions, leading to repeated questions, inconsistent work, and wasted time re-establishing context that was already shared.

Core Features & Use Cases

  • Hybrid Search Accuracy: Combines vector similarity and temporal knowledge graph results to return the most relevant memories, prioritizing graph-sourced results when conflicts arise for higher reliability.
  • Cross-Platform Compatibility: Works seamlessly in both Claude Code and Claude Cowork environments via the primary recall_memories MCP tool, with a secure REST fallback for edge cases where MCP is unavailable.
  • Flexible Scoping & Efficiency: Supports project-specific searches (with auto-detected project IDs) or global cross-project scans, plus an index-only mode for low-cost broad surveys of stored memories without loading full payloads.
  • Use Case: If you need to recall what technical stack choices you made for your current project 3 months ago, this skill retrieves matching memories with relevance scores and source attribution in seconds.

Quick Start

Ask the AI to search your Neuralscape memory for "what did we decide about the authentication system for the mobile app".

Frequently Asked Questions about search

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

FAQPage Schema
How do I recall past project decisions and preferences across AI assistant sessions?▼

To recall past project decisions across AI sessions, use a hybrid vector and temporal knowledge graph memory search to retrieve relevant stored memories with relevance scores and source attribution.

What is a hybrid memory search and how does it retrieve historical context?▼

A hybrid memory search combines vector similarity with a temporal knowledge graph to retrieve historical context, prioritizing graph-sourced results when conflicts arise to ensure higher reliability and accurate recall.

Can I search for specific memories within a single project environment?▼

Yes, you can search for specific memories within a single project environment using auto-detected project IDs for scoped filtering, or you can perform a global cross-project scan to find broader context.

Does memory search work in both Claude Code and Claude Cowork environments?▼

Memory search works seamlessly in both Claude Code and Claude Cowork environments via the primary recall_memories MCP tool, with a secure REST API fallback for edge cases where MCP tools are unavailable.

What is the best way to do a low-cost broad survey of stored memories?▼

The best way to do a low-cost broad survey of stored memories is to use the index-only mode, which scans the memory store without loading full payloads to reduce retrieval costs and overhead.

Related Skills

Explore More Software Engineering Skills→
SyntaxAsSpiralSyntaxAsSpiral
5

memory-systems

Persist AI agent context across sessions with vector stores and temporal knowledge graphs.

Community
Advanced
justlovemakijustlovemaki
79

prismflow-memory-read

Retrieve historical references from PrismFlow memory and knowledge bases.

Community
Intermediate
starrank-softstarrank-soft

autonomous-agent-harness

Integrate scheduled cron tasks, persistent memory, and computer-use capabilities into Claude Code.

Official
Advanced
heysuhasheysuhas
2

qmd

Index and search local markdown files with hybrid BM25 and vector retrieval.

Community
Advanced