/search

Search a Weaviate collection using hybrid, semantic, or keyword search

BOOST

weaviate@weaviate

GitHub

From plugin

weaviate

1049 skills9 commands

Install

> /plugin marketplace add weaviate/agent-skills
> /plugin install weaviate@weaviate-plugins

How it fires

How this command gets triggered: by you, by Claude, or both.

Fires itselfClaude auto-loads it when your prompt matches the work.
You can call itInvoke it directly when you want it.
Slash command/search

Context preview

What this command does when you run it.

Search a Weaviate collection using hybrid, semantic, or keyword search

Command definition

search.md

description: Search a Weaviate collection using hybrid, semantic, or keyword search
argument-hint: query [search text] collection [name] type [hybrid|semantic|keyword] alpha [0.5] limit [10]
allowed-tools: Bash, AskUserQuestion, Skill

Search Weaviate

Perform hybrid, semantic, or keyword search on a collection.

Usage

/weaviate:search query "your query" collection "CollectionName" [type "hybrid"] [alpha "0.5"] [limit "10"]

Workflow

When necessary, use AskUserQuestion to make entering arguments easier.

1. Parse arguments 2. If collection is missing:

Run `/weaviate:collections` to list available collections
Use AskUserQuestion to prompt user to select

3. If type is not specified, default to `hybrid` 4. Run the appropriate script based on type:

Hybrid (default)

uv run ${CLAUDE_PLUGIN_ROOT}/skills/weaviate/scripts/hybrid_search.py --query "USER_QUERY" --collection "COLLECTION_NAME" --alpha 0.7 --limit 10

Semantic

uv run ${CLAUDE_PLUGIN_ROOT}/skills/weaviate/scripts/semantic_search.py --query "USER_QUERY" --collection "COLLECTION_NAME" --limit 10

Keyword

uv run ${CLAUDE_PLUGIN_ROOT}/skills/weaviate/scripts/keyword_search.py --query "USER_QUERY" --collection "COLLECTION_NAME" --limit 10

Examples

/weaviate:search query "machine learning" collection "Articles"
/weaviate:search query "SKU-12345" collection "Products" type "keyword"
/weaviate:search query "similar concepts" collection "Documents" type "semantic"