skills.rest
sakamotomomotaro0809-netizen
sakamotomomotaro0809-netizenCOMMUNITY
Explore all skills by this author →
GitHub

qdrant-memory

Stores and retrieves text using Qdrant vector embeddings for semantic search.

Updated Mar 23, 2026
Search Tags:#semantic search#embeddings#rag#mcp#vector search#qdrant#long-term memory
One-click install
npx skills add https://github.com/sakamotomomotaro0809-netizen/tateyomi --skill qdrant-memory-sakamotomomotaro0809-netizen
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: qdrant-memory
Source: https://github.com/sakamotomomotaro0809-netizen/tateyomi/tree/main/taisun_agent/.claude/skills/qdrant-memory
Command: npx skills add https://github.com/sakamotomomotaro0809-netizen/tateyomi --skill qdrant-memory-sakamotomomotaro0809-netizen

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Keyword-based search only matches exact terms and misses semantically related content, making it hard to retrieve relevant knowledge across sessions. This Skill provides persistent vector-based memory so AI agents can store information and find conceptually similar content later. ## Core Features & Use Cases - Semantic Storage: Store text with vector embeddings (MiniLM-L6) in a Qdrant collection along with metadata like topic and date. - Similarity Search: Query stored memories by meaning rather than exact keywords, returning ranked similar documents. - Memory Management: Delete outdated entries by ID or metadata filter, and list available collections with stats. - Use Case: An AI agent stores a React state management pattern during one session, then in a later session retrieves it when asked about previously discussed state management approaches, using it as context for generating new code. ## Quick Start Start a local Qdrant server with Docker, then ask the agent to remember a piece of knowledge and later query it with a natural language question.

Frequently Asked Questions about qdrant-memory

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

FAQPage Schema
How do I set up Qdrant for vector search locally?▼

Run Qdrant locally with Docker using 'docker run -p 6333:6333 -v $(pwd)/qdrant_data:/qdrant/storage qdrant/qdrant', then verify it is running by calling http://localhost:6333/health. Set QDRANT_URL in your environment to point at the server.

How does semantic search differ from keyword search?▼

Semantic search matches content by meaning using vector embeddings, so it finds conceptually related documents even without exact keyword overlap. Keyword search only matches literal terms and misses paraphrased or related content.

Can I use Qdrant Cloud instead of a local server?▼

Yes, create a cluster at cloud.qdrant.io, then set QDRANT_URL to your cluster URL and QDRANT_API_KEY to your API key in the environment. The same store, find, and delete operations work against the cloud instance.

How do I delete outdated memories from Qdrant?▼

Use the qdrant-delete tool with an ID or a metadata filter, such as filtering by topic, to remove matching entries. Regular cleanup of outdated knowledge keeps search results relevant.

What metadata should I store with vector embeddings?▼

Store meaningful fields like topic, type, and date alongside each text entry, for example {"topic": "react", "type": "pattern", "date": "2026-01-19"}. This enables filtered retrieval and easier cleanup of stale entries.

Related Skills

Explore More Software Engineering Skills→
cdeistopenedcdeistopened
2

local-markdown-search

Index and search local Markdown collections with BM25 and vector embeddings.

Community
Intermediate
VictoriakaeyVictoriakaey

memory-system

Design memory systems for agents to persist and retrieve information across sessions.

Community
Advanced
starlink-awakenstarlink-awaken

missing-brain-first

Search external web indexes and synthesize results via external AI models.

Community
Basic
DbillionaerDbillionaer
1

vector-database-engineer

Engineer vector database solutions for semantic search and RAG applications.

Community
Advanced