Skip to content

Search Skills

Query smarter. Pre-built skills for documents, aggregations, and semantic search.

Quick Start

# 1. Install search skills
./hub/moltler-cli.sh install --all --category search
./hub/moltler-cli.sh install --all --category enterprise-search

# 2. Run your first skill
curl -u elastic-admin:elastic-password http://localhost:9200/_escript \
  -H "Content-Type: application/json" \
  -d '{"query": "RUN SKILL search_documents(query => '\''pricing'\'')"}'

Common Scenarios

"Find documents matching my query"

-- Simple search
RUN SKILL search_documents(query => 'pricing plans');

-- With filters
RUN SKILL search_documents(
  query => 'pricing',
  index_pattern => 'products-*',
  limit => 50
);

"What are the top values in this field?"

RUN SKILL top_values(
  index_pattern => 'orders-*',
  field => 'product_category',
  limit => 10
);

Returns aggregation of most common values.

"Search by meaning, not keywords"

-- Semantic/vector search
RUN SKILL semantic_search(
  query => 'how do I reset my password',
  index => 'support-articles'
);

Finds documents that match the meaning, even if words differ.

"Fuzzy match for typos"

RUN SKILL fuzzy_search(
  query => 'Jonh Smth',  -- typos
  index_pattern => 'customers-*',
  field => 'name'
);

"Show me a time histogram"

RUN SKILL date_histogram(
  index_pattern => 'orders-*',
  interval => 'day',
  time_range => '30d'
);

Available Skills

Document Operations

Skill Description
search_documents Full-text search
get_document Get document by ID
create_document Index a new document
update_document Update existing document
delete_document Remove a document
bulk_index Bulk index documents

Aggregations

Skill Description
top_values Terms aggregation
date_histogram Time-based bucketing
percentiles Percentile statistics
multi_field_search Search across fields
Skill Description
semantic_search Vector/embedding search
fuzzy_search Typo-tolerant search

Index Management

Skill Description
create_index Create new index
delete_index Remove index
get_mapping Get index mapping
set_alias Manage aliases
reindex Reindex data

Data Pipelines

Skill Description
list_ingest_pipelines List pipelines
test_ingest_pipeline Test pipeline
list_transforms List transforms
get_transform_status Transform health
Skill Description
list_search_apps Search applications
get_search_analytics Query analytics
get_top_queries Popular searches
get_no_results_queries Searches with no hits

Real-World Workflow

Scenario: Build a product search feature

-- Step 1: What fields do I have?
RUN SKILL get_mapping(index => 'products');
-- Result: name, description, category, price, rating

-- Step 2: What are the top categories?
RUN SKILL top_values(
  index_pattern => 'products',
  field => 'category',
  limit => 20
);
-- Result: Electronics, Clothing, Home, Sports...

-- Step 3: Test a search query
RUN SKILL search_documents(
  query => 'wireless headphones',
  index_pattern => 'products',
  limit => 5
);
-- Result: Matching products with scores

-- Step 4: Try semantic search for better results
RUN SKILL semantic_search(
  query => 'noise cancelling earbuds for running',
  index => 'products'
);
-- Result: Products matching intent, not just keywords

-- Step 5: Check search analytics
RUN SKILL get_no_results_queries(limit => 20);
-- Result: Queries users tried but found nothing

Data Pipeline Example

Scenario: Set up data ingestion

-- Step 1: Check existing pipelines
RUN SKILL list_ingest_pipelines();

-- Step 2: Test a pipeline
RUN SKILL test_ingest_pipeline(
  pipeline => 'logs-processor',
  document => '{"message": "ERROR: Connection failed"}'
);
-- Result: Processed document with extracted fields

-- Step 3: Bulk index data
RUN SKILL bulk_index(
  index => 'products',
  documents => [
    {"name": "Widget A", "price": 29.99},
    {"name": "Widget B", "price": 39.99}
  ]
);

-- Step 4: Verify the data
RUN SKILL search_documents(
  query => 'Widget',
  index_pattern => 'products'
);

Search Analytics

Understand how users search:

-- What are people searching for?
RUN SKILL get_top_queries(
  app => 'ecommerce-search',
  limit => 50
);

-- What searches return nothing?
RUN SKILL get_no_results_queries(
  app => 'ecommerce-search',
  limit => 20
);

-- Overall search health
RUN SKILL get_search_analytics(app => 'ecommerce-search');

Combine with AI

Let an AI assistant query your data:

User: "Find all products under $50 in the electronics category"

AI: [Runs search_documents with filters]

AI: "Found 47 products. Top results:
     1. Wireless Mouse - $24.99
     2. USB Hub - $19.99
     3. Phone Stand - $12.99

     Want me to filter further or show analytics?"

See MCP Integration to connect your AI assistant.

Build Custom Skills

Create search skills for your use case:

CREATE SKILL product_search
  VERSION '1.0.0'
  DESCRIPTION 'Search products with category and price filters'
  (
    query STRING DESCRIPTION 'Search query',
    category STRING DEFAULT NULL DESCRIPTION 'Filter by category',
    max_price NUMBER DEFAULT NULL DESCRIPTION 'Maximum price',
    limit INT DEFAULT 20 DESCRIPTION 'Results limit'
  )
  RETURNS ARRAY
BEGIN
  DECLARE base_query STRING;

  SET base_query = 'FROM products | WHERE MATCH(name, "' || query || '") OR MATCH(description, "' || query || '")';

  IF category IS NOT NULL THEN
    SET base_query = base_query || ' AND category == "' || category || '"';
  END IF;

  IF max_price IS NOT NULL THEN
    SET base_query = base_query || ' AND price <= ' || max_price;
  END IF;

  SET base_query = base_query || ' | SORT _score DESC | LIMIT ' || limit;

  RETURN ESQL_QUERY(base_query);
END SKILL;

Next Steps