Kurt

Kurt

Kurt is a content fetching and workflow automation tool designed for AI agents.

Quick Start

# Fetch a web page
kurt fetch https://example.com

# Discover URLs from sitemap
kurt map https://example.com/sitemap.xml

# List available tools and providers
kurt tool list

Automatic Provider Selection

Kurt automatically selects the best provider based on URL patterns:

Available Tools

fetch - Retrieve content from URLs

kurt fetch https://example.com                     # Auto-select provider
kurt fetch https://example.com --provider httpx     # Specific provider
kurt fetch https://example.com --json               # JSON output

map - Discover URLs from sources

kurt map https://example.com/sitemap.xml            # From sitemap
kurt map https://blog.example.com/feed.xml          # From RSS feed
kurt map ./docs --provider folder                    # From local folder
kurt map https://example.com --provider crawl        # Web crawl

workflow - Run automation workflows

kurt workflow list                                   # List workflows
kurt workflow run sync-docs.toml                     # Run workflow
kurt workflow run scrape-all.toml --background       # Background run

tool - Manage tools and providers

kurt tool list                                       # List all tools
kurt tool info fetch                                 # Tool details
kurt tool providers fetch                            # List providers
kurt tool check                                      # Check env requirements

Configuration

Kurt reads configuration from (highest to lowest priority):

1. CLI arguments
2. Project kurt.toml
3. User ~/.kurt/config.toml
4. Provider defaults

Example kurt.toml

[tool.fetch]
provider = "trafilatura"
timeout = 30

[tool.fetch.providers.notion]
include_children = true
max_depth = 3