Execution conventions
- CLI: deterministic operations go through
wewritecommand (must be in PATH; if missing, guide touv tool install wewrite, or in the repositorybash install.sh）。 - {home}: user state directory =
$WEWRITE_HOMEor~/.wewrite（wewrite homecan be checked). config/style/history/playbook/output/exemplars are all in {home}, not in the repository; state paths in the references documents follow this same convention. 读取: <路径>= actually finish reading the file with the file-reading tool before continuing, it is not a comment.- references/: this skill comes with
{skill_dir}/references/; the{skill_dir}inside the references documents is this skill's directory. - This module isinteractive's (it needs to ask the user questions), and is not bound by the main pipeline's "fully automatic" constraint.
Execute
Trigger scenarios：
- Main entry Step 1 finds that
{home}/style.yamldoes not exist (first-time use) - The user says "reset style", "modify configuration"
Both scenarios execute:
读取: {skill_dir}/references/onboard.md
Complete Phase 1-4 according to onboard.md: interactively collect information → generate {home}/style.yaml →
optional Playbook library setup → trial run inquiry. When resetting the style, use the existing style.yaml as the baseline, and only change the fields the user wants to change;
after changing, show the full text for the user to confirm.
After completion: if activated by the main pipeline (first-time onboard), return to the main pipeline Step 1 and continue; if triggered by the user separately, inform them "the configuration has been updated and will take effect automatically on the next writing task".