CLI Web Search

The Smart Search SDK provides powerful CLI commands for web searching, page fetching, and content extraction.

Overview

Web search commands allow you to:

Search the web for relevant documents and pages
Fetch web page URLs matching your query
Extract specific web page content
Get snippets from web pages based on queries
Extract keypoints from web pages

Prerequisites

Before using web search commands, ensure you have:

Smart Search SDK installed
SMARTSEARCH_BASE_URL environment variable set
Valid authentication credentials

Available Commands

Web Search

Search the web for documents or pages relevant to your query.

Usage

smart-search web search --query "<QUERY>" [OPTIONS]

Parameters

Parameter

Required

Type

Default

Description

--query

Yes

String

Search query string

--site

String

Comma-separated URLs to limit results to specific sites

--result-type

String

snippets

Type of results: snippets, keypoints, summary, description

--engine

String

auto

Search engine: auto, firecrawl, or perplexity

--size

Integer

5

Maximum number of results (1-50)

--stream

Boolean

false

Enable streaming response

--view

String

table

Output format: table or json

Examples

Basic search with snippets:

smart-search web search --query "Python tutorials"

Search with keypoints:

smart-search web search \
  --query "machine learning best practices" \
  --result-type keypoints \
  --size 10

Search with JSON output:

smart-search web search \
  --query "climate change solutions" \
  --view json \
  --size 20

Streaming search:

smart-search web search \
  --query "latest AI developments" \
  --stream true

Search with site filter:

smart-search web search \
  --query "machine learning frameworks" \
  --site "https://github.com" \
  --result-type snippets \
  --size 10

Search with multiple sites:

smart-search web search \
  --query "python frameworks" \
  --site "https://realpython.com,https://stackoverflow.com" \
  --result-type snippets \
  --size 10

Search with specific engine:

smart-search web search \
  --query "latest research papers" \
  --engine perplexity \
  --result-type keypoints \
  --size 15

Response (Table View)

┏━━━━┳━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━┓
┃ No ┃ URL                ┃ Title                ┃ Content Preview      ┃
┡━━━━╇━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━┩
│ 1  │ python.org/tut...  │ Python Tutorial      │ Learn Python step... │
│ 2  │ realpython.com/... │ Real Python Guide    │ Comprehensive Pyt... │
└────┴────────────────────┴──────────────────────┴──────────────────────┘
✓ Found 2 results.

Response (JSON View)

{
  "data": [
    {
      "content": "Learn Python step by step with examples...",
      "metadata": {
        "title": "Python Tutorial",
        "source": "https://docs.python.org/3/tutorial/"
      }
    }
  ]
}

Web - Map Website

Map a website and discover its URL structure and hierarchy.

Usage

smart-search web map --base-url "<BASE_URL>" [OPTIONS]

Parameters

Parameter

Required

Type

Default

Description

--base-url

Yes

String

Base URL of the website to map

--page

Integer

1

Page number to return (default is 1). Can be skipped if --return-all-map is true.

--size

Integer

20

Maximum number of URLs to return (1-1000), default is 20. Can be skipped if --return-all-map is true.

--return-all-map

Boolean

false

Return all mapped links (default is false)

--include-subdomains

Boolean

false

Include subdomains in the mapping

--query

String

Optional query to filter URLs by keywords

--view

String

table

Output format: table or json

Examples

Basic website mapping:

smart-search web map --base-url "https://docs.python.org"

Map with keyword filtering:

smart-search web map \
  --base-url "https://docs.python.org" \
  --query "tutorial" \
  --size 20 \
  --page 1

Include subdomains:

smart-search web map \
  --base-url "https://example.com" \
  --include-subdomains \
  --size 50

Return all mapped links:

smart-search web map \
  --base-url "https://docs.firecrawl.dev" \
  --return-all-map true

Get JSON output:

smart-search web map \
  --base-url "https://docs.firecrawl.dev" \
  --size 30 \
  --page 1 \
  --view json

Response (Table View)

🗺️  Mapping website...
Base URL: https://docs.python.org
Page: 1
Size: 20
Return All Map: False
Include Subdomains: False
Query: None
View: table

┏━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ #  ┃ URL                               ┃ Title                        ┃
┡━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
│ 1  │ https://docs.python.org/3/        │ Python 3 Documentation       │
│ 2  │ https://docs.python.org/3/tuto... │ The Python Tutorial          │
│ 3  │ https://docs.python.org/3/lib...  │ The Python Standard Library  │
└────┴───────────────────────────────────┴──────────────────────────────┘
✓ Found 20 URLs.

Response (JSON View)

{
  "data": [
    {
      "url": "https://docs.python.org/3/",
      "title": "Python 3 Documentation",
      "description": "Official Python documentation"
    },
    {
      "url": "https://docs.python.org/3/tutorial/",
      "title": "The Python Tutorial",
      "description": "Learn Python step by step"
    }
  ],
  "total_found_links": 42,
  "page": 1,
  "size": 20,
  "total_pages": 3
}

Web - Search URLs

Search the web for pages and return their URLs (similar to Google search results).

Usage

smart-search web urls --query "<QUERY>" [OPTIONS]

Parameters

Parameter

Required

Type

Default

Description

--query

Yes

String

Search query string

--site

String

Comma-separated URLs to limit results to specific sites

--engine

String

auto

Search engine: auto, firecrawl, or perplexity

--size

Integer

5

Maximum number of URLs (1-50)

--stream

Boolean

false

Enable streaming response

--view

String

table

Output format: table or json

Examples

Find URLs about Python documentation:

smart-search web urls --query "Python documentation" --size 10

Get URLs with JSON output:

smart-search web urls \
  --query "react best practices" \
  --view json

Search URLs with site filter:

smart-search web urls \
  --query "python tutorials" \
  --site "https://realpython.com" \
  --size 10

Search multiple sites for URLs:

smart-search web urls \
  --query "ai tutorials" \
  --site "https://medium.com,https://dev.to" \
  --size 20

Search with specific engine:

smart-search web urls \
  --query "cloud computing guides" \
  --engine firecrawl \
  --size 10

Response (Table View)

┏━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
┃ No ┃ URL                                      ┃
┡━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
│ 1  │ https://docs.python.org/3/              │
│ 2  │ https://www.python.org/doc/             │
│ 3  │ https://realpython.com/                 │
└────┴──────────────────────────────────────────┘
✓ Found 3 URLs.

Web - Fetch Page

Fetch a single web page by URL and return its content and metadata.

Usage

smart-search web fetch --url "<URL>" [OPTIONS]

Parameters

Parameter

Required

Type

Default

Description

--url

Yes

String

URL of the web page to fetch

--json-schema

String

Path to JSON schema file or JSON string

--return-html

Boolean

false

Return full HTML content

--stream

Boolean

false

Enable streaming response

--view

String

json

Output format: json or markdown

Examples

Fetch page content:

smart-search web fetch --url "https://docs.python.org/3/tutorial/"

Fetch with HTML:

smart-search web fetch \
  --url "https://example.com/article" \
  --return-html \
  --view markdown

Fetch with JSON schema:

smart-search web fetch \
  --url "https://example.com/product-page" \
  --json-schema '{"type": "object", "properties": {"product_name": {"type": "string"}, "price": {"type": "number"}, "features": {"type": "array", "items": {"type": "string"}}}}' \
  --view json

Fetch with schema from file:

smart-search web fetch \
  --url "https://example.com/article" \
  --json-schema "@schema.json" \
  --view json

Response (Markdown View)

✓ Page fetched successfully
Title: Python Tutorial
Description: The Python Tutorial — Python 3.x documentation
Language: en
Status: 200

Page Content (Markdown):

# Python Tutorial

Python is an easy to learn, powerful programming language...

Response (JSON View)

{
  "metadata": {
    "title": "Python Tutorial",
    "description": "The Python Tutorial",
    "language": "en",
    "status_code": 200
  },
  "markdown": "# Python Tutorial\n\nPython is...",
  "html": "<html>...</html>"
}

Web - Get Web Page Snippets

Extract relevant text snippets from a web page for a given query.

Usage

smart-search web snippets --query "<QUERY>" --url "<URL>" [OPTIONS]

Parameters

Parameter

Required

Type

Default

Description

--query

Yes

String

Query to search within the page

--url

Yes

String

URL of the web page to extract from

--size

Integer

5

Maximum number of snippets (1-50)

--json-schema

String

Path to JSON schema file or JSON string

--style

String

paragraph

Snippet style: paragraph or sentence

--stream

Boolean

false

Enable streaming response

--view

String

table

Output format: table or json

Examples

Extract snippets about list comprehension:

smart-search web snippets \
  --query "list comprehension" \
  --url "https://docs.python.org/3/tutorial/" \
  --size 3

Extract sentence-style snippets:

smart-search web snippets \
  --query "async programming" \
  --url "https://docs.python.org/3/library/asyncio.html" \
  --style sentence \
  --size 5

Extract with JSON schema:

smart-search web snippets \
  --query "python examples" \
  --url "https://realpython.com/python-decorators/" \
  --json-schema '{"type": "object", "properties": {"code_examples": {"type": "array", "items": {"type": "object", "properties": {"language": {"type": "string"}, "code": {"type": "string"}}}}}}' \
  --size 5 \
  --view json

Response (Table View)

┏━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━┓
┃ No ┃ Content                         ┃ Score  ┃
┡━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━┩
│ 1  │ List comprehensions provide...  │ 0.92   │
│ 2  │ A list comprehension consists...│ 0.88   │
└────┴─────────────────────────────────┴────────┘
✓ Found 2 snippets.

Web - Get Web Page Keypoints

Extract concise key points from a web page.

Usage

smart-search web keypoints --query "<QUERY>" --url "<URL>" [OPTIONS]

Parameters

Parameter

Required

Type

Default

Description

--query

Yes

String

Query to search for

--url

Yes

String

URL of the web page to extract from

--size

Integer

5

Maximum number of keypoints (1-50)

--json-schema

String

Path to JSON schema file or JSON string

--stream

Boolean

false

Enable streaming response

--view

String

table

Output format: table or json

Examples

Extract Python features:

smart-search web keypoints \
  --query "Python features" \
  --url "https://www.python.org/about/" \
  --size 5

Get keypoints with JSON output:

smart-search web keypoints \
  --query "benefits of TypeScript" \
  --url "https://www.typescriptlang.org/" \
  --view json

Extract keypoints with JSON schema:

smart-search web keypoints \
  --query "product information" \
  --url "https://example.com/product" \
  --json-schema '{"type": "object", "properties": {"features": {"type": "array", "items": {"type": "string"}}, "benefits": {"type": "array", "items": {"type": "string"}}}}' \
  --size 5 \
  --view json

Response (Table View)

┏━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━┓
┃ No ┃ Key Point                       ┃ Score  ┃
┡━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━┩
│ 1  │ Python is easy to learn         │ 0.95   │
│ 2  │ Extensive standard library      │ 0.91   │
│ 3  │ Dynamic typing system           │ 0.87   │
└────┴─────────────────────────────────┴────────┘
✓ Found 3 key points.

Output Formats

Table View

Human-readable format
Best for interactive CLI usage
Automatically formats columns and rows

JSON View

Machine-readable format
Best for scripting and automation
Complete data structure with all fields

Streaming Responses

Enable streaming for better performance with large result sets:

smart-search web search --query "AI news" --stream true

Benefits:

Real-time results as they arrive
Reduced memory usage
Better user experience for long-running queries

Best Practices

Choose appropriate result types:
- Use snippets for detailed text excerpts
- Use keypoints for concise summaries
- Use summary for high-level overviews
Optimize size parameter:
- Start with default (5) for quick results
- Increase for comprehensive searches
- Maximum is 50 results per query
Use streaming for performance:
- Enable --stream true for queries returning many results
- Reduces waiting time and memory usage
Select appropriate snippet style:
- paragraph: For contextual information
- sentence: For precise, concise answers
JSON for automation:
- Use --view json when integrating with scripts
- Easier to parse and process programmatically

Common Use Cases

Research and Documentation

# Find comprehensive guides
smart-search web search \
  --query "TypeScript advanced patterns" \
  --result-type snippets \
  --size 10

# Extract specific information
smart-search web snippets \
  --query "dependency injection" \
  --url "https://angular.io/guide/dependency-injection"

Quick Information Lookup

# Get concise answers
smart-search web keypoints \
  --query "React hooks useEffect" \
  --url "https://react.dev/reference/react/useEffect"

Content Aggregation

# Collect URLs for further processing
smart-search web urls \
  --query "climate change research papers 2024" \
  --size 50 \
  --view json > urls.json

Page Analysis

# Analyze webpage content
smart-search web fetch \
  --url "https://example.com/article" \
  --view markdown > article.md

Error Handling

All commands provide clear error messages:

❌ Error occurred during web search
Details: Invalid URL format - please provide a valid HTTP/HTTPS URL

Common errors:

Invalid URL: Ensure URLs start with http:// or https://
Connection timeout: Check network connectivity and SMARTSEARCH_BASE_URL
Authentication failed: Verify credentials and environment variables
Size out of range: Use values between 1 and 50

PreviousCLI Commands NextCLI Connector

Last updated 1 month ago

hashtagOverview

hashtagPrerequisites

hashtagAvailable Commands

hashtagWeb Search

hashtagUsage

hashtagParameters

hashtagExamples

hashtagResponse (Table View)

hashtagResponse (JSON View)

hashtagWeb - Map Website

hashtagUsage

hashtagParameters

hashtagExamples

hashtagResponse (Table View)

hashtagResponse (JSON View)

hashtagWeb - Search URLs

hashtagUsage

hashtagParameters

hashtagExamples

hashtagResponse (Table View)

hashtagWeb - Fetch Page

hashtagUsage

hashtagParameters

hashtagExamples

hashtagResponse (Markdown View)

hashtagResponse (JSON View)

hashtagWeb - Get Web Page Snippets

hashtagUsage

hashtagParameters

hashtagExamples

hashtagResponse (Table View)

hashtagWeb - Get Web Page Keypoints

hashtagUsage

hashtagParameters

hashtagExamples

hashtagResponse (Table View)

hashtagOutput Formats

hashtagTable View

hashtagJSON View

hashtagStreaming Responses

hashtagBest Practices

hashtagCommon Use Cases

hashtagResearch and Documentation

hashtagQuick Information Lookup

hashtagContent Aggregation

hashtagPage Analysis

hashtagError Handling

Overview

Prerequisites

Available Commands

Web Search

Usage

Parameters

Examples

Response (Table View)

Response (JSON View)

Web - Map Website

Usage

Parameters

Examples

Response (Table View)

Response (JSON View)

Web - Search URLs

Usage

Parameters

Examples

Response (Table View)

Web - Fetch Page

Usage

Parameters

Examples

Response (Markdown View)

Response (JSON View)

Web - Get Web Page Snippets

Usage

Parameters

Examples

Response (Table View)

Web - Get Web Page Keypoints

Usage

Parameters

Examples

Response (Table View)

Output Formats

Table View

JSON View

Streaming Responses

Best Practices

Common Use Cases

Research and Documentation

Quick Information Lookup

Content Aggregation

Page Analysis

Error Handling