AI-assisted learning workflows for homeschool reporting.

This project uses Poetry for dependency management.

poetry install

For common operations, prefer the Taskfile:

task --list

Run the full test suite:

task test

Run the reporting-focused tests:

task test-reporting

Generate the current week's summary:

task summarize-current

Regenerate the current week's summary:

task summarize-current-force

Generate the current week's summary with additional context:

task summarize-current COMMENT="Student was sick this week"

If the weekly text summary is already correct and only the scans PDF needs to be refreshed, use the dedicated scans rebuild workflow.

Rebuild the current week's scans PDF:

task rebuild-scans-current-force

Rebuild the previous week's scans PDF:

task rebuild-scans-previous-force

Rebuild a week by offset from the current week:

task rebuild-scans-week OFFSET=2 EXTRA_ARGS="--force"

You can also run the script directly:

poetry run python scripts/rebuild_week_scans.py --current --force

Process Nomy PDF reports:

task process-nomy

Parse Sofatutor exports into activity JSON:

task update-sofatutor

Build an activity report grouped by subject:

task activities-by-subject START=2026-01-01 END=2026-01-31

• Weekly scans support Gescannt_*, Pixel PXL_*, and WhatsApp IMG-YYYYMMDD-WAxxxx.* filenames.
• Pixel and WhatsApp images are normalized into Gescannt_* filenames during scan PDF rebuilds.