跳转至

course-directory-structure

所属包: course

Create, reorganize, normalize, or audit a course project directory tree,

兼容性: Designed for OpenCode. Assumes repo-local .opencode/skills discovery

Purpose

Create and maintain a predictable course project directory structure for OpenCode-based course development.

When to use

Activate when the user asks to:

  • initialize a new course repository
  • create a standard course artifact tree
  • reorganize messy course files
  • normalize naming and placement
  • archive old drafts
  • check whether a course project is structurally complete

Default directory model

Use this structure unless the user provides a better existing standard:

.opencode/
  skills/
docs/
  00-project/
  01-outline/
  02-content/
  03-labs/
  04-learner-pack/
  05-instructor-pack/
  06-qa/
  07-qc/
assets/
  images/
  drawio/
  tables/
references/
  external/
  internal/
release/
archive/

Naming rules

  • Prefer lowercase kebab-case for file names.
  • Prefix ordered directories and ordered lesson files with two digits.
  • Keep one primary topic per file.
  • Separate learner-facing and instructor-only files.
  • Keep generated reports under docs/06-qa/ or docs/07-qc/, not mixed into content.
  • Move obsolete or superseded drafts into archive/, not random old/ folders.

Operating procedure

  1. Inspect the current tree.
  2. Infer the user's intended course lifecycle.
  3. Compare the current tree with the default model.
  4. Propose the minimal safe reorganization.
  5. If the repository is messy, split the work into three phases: initialize -> audit -> reorganize.
  6. Create missing folders and move files only when the target location is unambiguous.
  7. When ambiguity exists, generate a dry-run style reorganization proposal instead of guessing.
  8. Record structural decisions in a project note if the reorganization is substantial.

Use the scaffold script when appropriate

Script: uv run scripts/bootstrap_course_tree.py --help

Use it when the user explicitly wants a new scaffold or a normalized tree created on disk.

Use the audit script when appropriate

Script: uv run scripts/check_course_tree.py --help

Use it when the user asks whether the repository is complete, clean, or ready for the next phase.

Output expectations

When reviewing a structure, report:

... (完整 SKILL.md 中还有 20 行)