Editor’s note: This article is Part 3 of a three-part series on AI-assisted course design in Brightspace. In Part 1, Kumar explored why generic AI tools often struggle with platform-specific course design workflows. In Part 2, he introduced the Brightspace Interactive Design Skill Suite and its architecture. If you’re new to the series, start with Part 1: The Gap No One Is Talking About: Why Generic AI Falls Short in Brightspace and follow with Part 2: Meet the Brightspace Interactive Design Skill Suite before diving into this part.
The walkthrough below is drawn from a real course build using the suite. Dialogue has been lightly edited for clarity and institutional references genericized.
The GitHub repository is now public. The suite is ready to use. This article covers what you need to get from nothing to a running first session — including the things most setup guides skip: what happens to your data, what your copyright obligations are and what the experience actually feels like.
What You Need
You need a Claude account, a GitHub account to download the skill files and a browser. Course materials help but aren’t required. The suite works from a blank slate. Please note that the free Claude tier doesn’t support Projects. A paid account is required. If you are on a consumer plan (Free, Pro, or Max), check whether your conversations are set to contribute to model training before your first session. See the Privacy section below for steps.
Setup: Skills and Projects
Setup takes about 10 minutes across two stages. Steps 1 to 3 are account-level and done once. Steps 4 and 5 are course-level and repeated for each new course.
| 01 | Enable Code execution in Settings Go to claude.ai/settings/capabilities and make sure the Code execution and file creation toggle is switched on. This is a one-time account-level setting required before skills will appear or run. |
| 02 | Download the skill files from GitHub Go to the GitHub repository (link in the Get Started section). Click the green Code button, then Download ZIP. Unzip it. You will find 41 subfolders, one per skill. Each subfolder contains a single SKILL.md file. That’s all you need. |
| 03 | Upload each skill via Customize Skills Go to Settings > Capabilities. Under the Skills section, click Customize. For each of the 41 skills: click + > Create skill > Upload a skill > open the skill’s subfolder and select the SKILL.md file. Toggle each skill on after upload. Skills are account-level— upload once and they are available everywhere. Upload brightspace orchestrator last so it’s easy to find. |
| 04 | Create a Project for your course In the left sidebar, click Projects > New Project. Give it a name and a short description — for example, name: GNED1101: Course Design, description: Firstyear course design workspace. Create one Project per course. Skills are already available — you don’t upload them again here. |
| 05 | Paste your Project Instructions when prompted in your first session Project Instructions stores your course context — course name, delivery mode, brand colours, term length, and pointers to your knowledge files. You don’t write this yourself. During your first session, the orchestrator collects this information and gives you a ready-to-paste block. When it does, open your Project, click Edit Project Instructions, and paste it in. This happens once per course. |
| 06 | Review your privacy settings Go to claude.ai/settings/data-privacy-controls and review the ‘Help improve Claude’ toggle. Switch it off if you prefer your conversations not to be used for model training. Check your institution’s guidance if unsure. This is your decision. |
| 07 | Start your first session with /brightspace-orchestrator Click New Chat inside your course Project. Type / and a list of your available skills appears. Select brightspace-orchestrator. Then describe your course design need in plain language. The orchestrator routes you automatically. You invoke the orchestrator once per session. Everything after that is handled automatically. |
Suite.
Your First Session
Example 1— Starting from scratch
What follows is a real first session. An instructor building a first-year Introduction to Cell and Molecular Biology course from scratch — no outline, no materials, no prior course to copy from.
Faculty: /brightspace-orchestrator I’m building a new course for the first time and I have nothing yet. Where do I start?
Orchestrator: Great. This is clearly a full course design from scratch. A few quick questions before routing you to the right starting point.
Q: Do you have anything to start with? Even rough notes, a syllabus from a similar course, an institutional template, or a list of topics?
Faculty: Nothing at all. Completely blank slate.
Q: What’s the delivery mode for this course?
Faculty: Blended / hybrid.
Q: Are you working inside a Claude Project for this course?
Faculty: No, but I’d like to set one up.
Orchestrator: Perfect. Before I route you, two more quick ones.
Q: What’s the course code and name?
Faculty: Still being decided.
Q: Do you have institutional brand colours, or should I use a standard palette?
Faculty: I have custom colours to specify.
Orchestrator: Go ahead and share two things when you’re ready: your custom colours (hex codes if you have them, or just describe the palette), and what the course is about. Even a rough description like ‘introductory stats for business students’ is enough to get started.
[Instructor shares a colour palette image and describes the course]
Orchestrator: Got it. One last question before we set up your Project.
Q: How long is the term, and roughly how many contact hours per week?
Faculty: 13 weeks, 3 hrs/week.
Orchestrator: Perfect. I have everything I need. Here’s everything you need to get started.
Step 1: Create your Claude Project In Claude.ai, click Projects in the left sidebar, then + New Project. Name it after your course.
Suggested Title: Cell and Molecular Biology— Intro (First Year)
Step 2: Paste this into Project Instructions (Fill in the brackets with your details. Replace [Your Institution] with your institution’s name.)
I am designing Introduction to Cell and Molecular Biology at [Your Institution].
Term: [TBD — update when confirmed]
Delivery mode: Blended / Hybrid
Contact hours: 3 hrs/week · 13 weeks
Enrolment: [update when known]
Student level: First year
Brand colours:
Primary (Gold): #946E1C
Secondary (Navy): #263D69
Core Blue: #1C4694
Dark Brown: #614E26
Cream: #FFEEC9
Sky Blue: #B0CCFF
Dark Navy: #212B3F
Warm Yellow: #FADA93
Soft Cream: #FFF0CF
Pale Blue: #DAE7FF
Current phase: Inception
My course outline is in the knowledge files as course-outline.md
My session log is in the knowledge files as session-log.md.
My course design status is in the knowledge files as course-design-status.md.
Read these files at the start of every session before responding.
I am using the Brightspace Interactive Design Skill Suite.
Always check the knowledge files for current course context before asking setup questions.
Step 3: Add two starter files to the Project the orchestrator generates a session-log.md and course-design-status.md. Download both and upload them to your Project via Add content, then Upload files. Name them exactly as shown.
Step 4: Your first skill:
/brightspace-course-outline-builder I’m building a first-year Introduction to Cell and Molecular Biology course from scratch. Blended delivery, 13 weeks, 3 hrs/week. No outline yet — please run Workflow 1.
The outline builder will interview you about topics, assessments, prerequisites, and course structure, then produce a complete course-outline.md to add to the Project. Every skill after that reads from it automatically. Your full recommended build sequence is saved in course-design-status.md.
Orchestrator: Ready to set up the Project, or do you have questions before you go?
The orchestrator detected a cold-start scenario, collected five intake items, read the color palette from an uploaded image and produced a ready-to-paste Project instructions block with the full build sequence. Brand colours, term length, and delivery mode are preloaded for every skill that follows. The instructor sets this up once.
Example 2— Redesigning from existing materials
An instructor redesigning a course they have taught before. A syllabus and slides are ready to upload.
Faculty: /brightspace-orchestrator I’m redesigning a course I’ve taught before. I have a syllabus and some slides. Where do I start?
Orchestrator: Good. This is clearly a full course redesign. A few quick questions before I route you.
Q: Are you working inside a Claude Project set up for this course?
Faculty: No, I’m working in a regular chat.
Q: What’s your primary goal for this redesign?
Faculty: All of the above. Full rebuild.
Q: What do you have ready to upload?
Faculty: Syllabus / course outline (Word or PDF) and lecture slides (PPTX).
Orchestrator: Good. That’s enough to route you clearly. For a full rebuild, setting up a Claude Project first will save significant time. Without it, every new conversation loses your course context and you’ll re-explain yourself repeatedly. With a Project, every skill session starts with full context automatically. It takes about two minutes.
Q: What’s the course code and delivery mode?
Faculty: UGST 1001. Hybrid / blended.
Q: What’s the course name, as it appears on your syllabus?
Faculty: It’s in the syllabus. Just parse it when I upload.
Q: How many weeks is the term?
Faculty: 13 weeks.
Orchestrator: Perfect. Here’s your ready-to-paste Project Instructions block. Replace [Your Institution] with your institution’s name and add your brand colors if you have them.
I am designing UGST 1001 at [Your Institution].
Delivery mode: Hybrid / Blended
Brand colours: [add your institution’s hex codes here]
Term length: 13 weeks
Current phase: Inception
My course outline is in the knowledge files as course-outline.md.
My session log is in the knowledge files as session-log.md.
My course design status is in the knowledge files as course-design-status.md.
Read these files at the start of every session before responding.
I am using the Brightspace Interactive Design Skill Suite.
Always check the knowledge files for current course context before asking setup questions.
Orchestrator: When you’re ready to start building:
- Set up the Project and paste the block above into Project Instructions.
- Open a conversation inside that Project.
- Type /brightspace-course-outline-builder and upload your syllabus.
Orchestrator: That first skill will parse the syllabus, extract your outcomes, assessment structure, and schedule, and produce a clean course-outline.md you save to the Project knowledge files. Everything downstream builds from that. Upload your slides at the same time if you like. I’ll note them for the slide-converter step once the outline is done.
Your recommended build sequence after that:
- course-outline-builder: parse syllabus into a clean foundation document
- content-currency-auditor: flag outdated content in your existing materials
- learning-outcomes-generator: verify and refine course-level outcomes
- redesign-planner: decide which modules get HTML Topics, what gets rebuilt
- gradebook planner: set up categories and weights before any assessments
- html-builder + slide-converter: build interactive content
- quiz / assessment / discussion generators: build assessments
- accessibility-auditor + student-experience-preview: QA before launch
Orchestrator: Ready when you are.
The orchestrator identified a redesign scenario, caught that no Project was set up, collected intake in five questions, and produced a ready-to-paste instructions block with a full eight-step build sequence. When the instructor said ‘just parse it when I upload,’ the orchestrator accepted that without friction — it adapted to what the instructor wanted to provide, not what it expected to receive.
What the Experience Actually Feels Like
One invocation starts everything
At the start of each session, type / in your Project chat and select brightspace-orchestrator from the skill list. That is the only skill you call manually. Describe what you need in plain language. The orchestrator reads your intent and routes to the right specialist — or asks one clarifying question if the intent is ambiguous. The other 40 skills run in the background, called by the orchestrator as needed. You never invoke them directly.
Context persists across sessions
After your first session, the Course Outline Builder will prompt you to save courseoutline.md to your Project knowledge base. Every skill that runs after that reads the file at session start. You never re-explain your course.
Your expertise is the input
The suite makes no pedagogical decisions. It handles sequencing, platform knowledge, and prerequisite checking. The instructional expertise stays yours.
For LMS Administrators and Instructional Designers
Customizing the suite for your institution
Skill files are plain Markdown, editable in any text editor. Common edits include adding institutional branding to HTML Topic output, incorporating assessment D2L Blog Series · Part 3 of 3 Kumar Chandrasekhar, PhD · Mount Royal University · CC BY-NC 4.0 · 10 policies specific to your LMS, adjusting department terminology and building new specialist skills for workflows the public suite does not cover. Custom skills upload to Customize > Skills alongside the public ones and work identically.
Distributing the suite to faculty
Skills are account-level. Once uploaded via Customize Skills, they are available in any chat. Institutions distributing the suite to faculty can point them to the README quickstart guide, which walks through the upload process step by step. Faculty create a Project for each course, paste in the instructions block the orchestrator generates and start a session.
Building institution-specific skills
The public GitHub repository contains the 40 skills in this series. Institutions can build additional specialist skills for their own workflows and upload them alongside. The README covers SKILL.md structure and the metadata format the orchestrator uses for routing. New skills slot in without touching any existing files.
Privacy and Data Handling
What data goes to Anthropic
When you use Claude, conversation content — including anything you type or paste — goes to Anthropic’s servers for processing. Data is encrypted in transit (TLS) and at rest (AES-256). Anthropic does not sell user data to third parties.
Training data opt-out
For consumer accounts (Free, Pro and Max), Anthropic may use your conversations to train future models if you choose to allow it. Training is opt-in — you must make an active choice. If you opt in, conversations are retained in de-identified form for up to five years. If you opt out, the standard 30-day retention applies and your data is not used for training. Team, Enterprise, and institutional accounts are not subject to these training practices. Steps: (1) Go to Settings > Privacy > Privacy Settings (2) Find the ‘Help improve Claude’ toggle (3) Switch off if preferred. Takes effect immediately. Check with your institution’s IT team if unsure what account type applies.
Some account arrangements offer Zero Data Retention (ZDR), where conversation data is not stored beyond immediate processing. This is the right configuration for institutions with FERPA, PIPEDA, or GDPR obligations. Talk to your institution’s IT team or Claude account representative.
FERPA considerations (US institutions)
Do not paste student names, student IDs, grades, or individually identifiable student information into Claude sessions. Course design content — outlines, outcomes, assessment design — contains no student data and is generally FERPAcompliant. For specific use cases, consult your institution’s privacy officer.
PIPEDA and provincial privacy law (Canadian institutions)
Do not input student-identifiable information. Course design content is generally compliant. Institutions in provinces with data residency requirements (Alberta PIPA, BC PIPA) should confirm with their privacy office whether cloud AI processing outside Canada is permissible. Claude’s servers are primarily US-based.
Simple rule: use the suite for course design; outlines, outcomes, assessment design, content creation. Never input student-identifiable data. When in doubt, ask your institution’s privacy office.
Third-party AI Tools (Flow and NotebookLM Skills)
Two skills in the suite, google-flow-video-pipeline and notebooklm-video-builder, produce videos using Google services rather than Claude. When you use them, your prompts and the sources you provide go to Google under Google’s terms, not Anthropic’s. NotebookLM in particular builds video from documents you upload, so use only materials you have the rights to. Everything in the sections above still applies: keep student-identifiable data out, and bring only rights-cleared content into any session.
Copyright
The suite — CC BY-NC 4.0
The Brightspace Interactive Design Skill Suite is licensed under Creative Commons Attribution-NonCommercial 4.0. You may use, share, and adapt it for any noncommercial educational purposes with attribution. Commercial use requires permission.
Attribution: Brightspace Interactive Design Skill Suite by Kumar Chandrasekhar, PhD. Licensed under CC BY-NC 4.0.
| You can | You cannot |
| Use freely for course design at your institution | Sell or offer as a paid service |
| Adapt skill files for your institution | Remove author credit from distributed versions |
| Share with colleagues across institutions | Use in a commercial instructional design product |
| Contribute improvements to the public repository | Sublicence under more restrictive terms |
Your course content
Content you create using the suite is yours. The suite is a design tool; the intellectual property of the output belongs to you.
Third-party materials you bring into sessions
Do not paste full copyrighted readings, textbook excerpts or third-party content you do not have rights to. Use bibliographic information instead. The Reading List Builder works from titles and descriptions, not pasted text.
AI-generated content and copyright
Most copyright offices currently take the position that purely AI-generated content without meaningful human input may not qualify for copyright protection. The suite involves substantial human direction — your course design expertise is the creative input. Treat the output the same way you would treat a document you wrote with significant editorial help.
Get Started
The repository is public. Everything you need — skill files, README, faculty quickstart guide, and installation instructions — is there.
License: CC BY-NC 4.0 | Author: Kumar Chandrasekhar, PhD
If you run into a configuration issue or unexpected skill behavior, the repository Issues tab is the right place to report it. You need a free GitHub account to post — sign up at github.com if you don’t have one. Before posting, search the existing issues in case someone has already reported the same thing and received a fix. When you do post, include which skill triggered the problem, what you asked for and what happened instead. Feature requests and skill contributions are welcome there too.
This post is Part 3 of a three-part series on AI-assisted course design in Brightspace. The Brightspace Interactive Design Skill Suite is licensed under CC BY-NC 4.0.
Author’s note: This post was drafted with the assistance of an AI writing tool. All ideas, arguments, examples and conclusions are my own. I reviewed, revised and take full responsibility for the content as published.
Written by:
