nightshift:archive

Archive shift $ARGUMENTS.

Steps

1. Resolve the shift

   • If $ARGUMENTS is non-empty, use it.
   • Otherwise, list directories in .nightshift/ (excluding archive/); auto-select if exactly one exists, otherwise use AskUserQuestion to pick.
   • If no active shifts exist, report "No active shifts to archive." and stop.

2. Validate the shift exists

   Check .nightshift/<shift>/ exists. If not, report the error and stop.

3. Check for incomplete items

   For each task column (from flock -x .nightshift/<shift>/table.csv qsv headers --just-names .nightshift/<shift>/table.csv):

   flock -x .nightshift/<shift>/table.csv qsv search --exact done --select <task-column> --invert-match .nightshift/<shift>/table.csv | qsv count
   

   If any non-done items exist, break down by status:

   flock -x .nightshift/<shift>/table.csv qsv search --exact todo --select <task-column> .nightshift/<shift>/table.csv | qsv count
   flock -x .nightshift/<shift>/table.csv qsv search --exact failed --select <task-column> .nightshift/<shift>/table.csv | qsv count
   

   Warn the user with the breakdown and use AskUserQuestion to confirm:

   "Archive anyway?"

   Options: "Yes, archive with incomplete items" / "No, cancel". If the user cancels, stop.

4. Run the bundled archive script

   ${CLAUDE_SKILL_DIR}/scripts/archive.sh <shift>
   

   The script computes today's date in ISO format and moves the directory atomically. It exits non-zero on collision (.nightshift/archive/YYYY-MM-DD-<shift>/ already exists).

5. Confirm

   ## Shift Archived
   
   **Shift:** <shift>
   **Location:** `.nightshift/archive/YYYY-MM-DD-<shift>/`
   
   All files preserved (manager.md, table.csv, task files).
   

Guardrails

• Always use today's date in ISO YYYY-MM-DD format for the archive prefix.
• Never overwrite an existing archive.
• Warn about incomplete items but allow archiving with explicit confirmation.
• Preserve all files in the move (manager.md, table.csv, all task .md files).