[kirkwat]

• jobspy [QUERY] [options] runs scrape_jobs from the shell; the options mirror its parameters, and jobspy -h lists them, every output field and the exit codes
• One JSON object per job on stdout (JSON Lines), 5 jobs from each of the four general boards and the main fields by default; --fields picks fields, --fields all prints every one; -o FILE also saves the full records, as JSON Lines, a JSON array (.json) or CSV (.csv) by file name
• Errors and warnings go to stderr; exit 0 when at least one board answered (also with no jobs), 1 when every board failed, 2 for a usage error
• Installed as a console script ([project.scripts]) and runnable as python -m jobspy; no new dependency
• __all__ lists scrape_jobs as the public import; from jobspy import * no longer pulls in the scraper classes