version: "1.0.0" description: "Manage data sources for API, RSS, web scraping, files, and databases" disable-model-invocation: false allowedTools: mcp__krowork__krowork_register_datasource, mcp__krowork__krowork_list_datasources, mcp__krowork__krowork_fetch_datasource, mcp__krowork__krowork_delete_datasource

KroWork: Data Sources

Register, manage, and fetch data from external data sources.

Workflow

Step 1: Understand Data Need

The user input $ARGUMENTS describes the data source they want to set up. Determine:

• Source type: REST API, RSS feed, web scraping, local file, or SQLite database
• Configuration: URL, authentication, headers, query parameters
• Purpose: One-time fetch or repeated use

Step 2: Register Data Source

If the user wants a persistent data source, use krowork_register_datasource:

• name: Human-readable name (e.g., "Hacker News")
• source_type: One of: rest_api, rss, web_scrape, local_file, sqlite
• config: Source-specific configuration

Step 3: Fetch Data

Use krowork_fetch_datasource with the registered name to retrieve data.

Step 4: Present Results

Show the fetched data in a readable format and suggest how to use it:

• Create an app with this data source
• Set up monitoring/scheduling
• Export the data

Supported Source Types