What these skills are
One course, from the request to the release, as installable Claude skills.
This pack is for instructional designers, learning experience designers, e-learning developers and the L&D person who builds courses alone.
The skills follow one course in the order you build it, in eight stages, from the brief that restates the request as a problem to the sign-off record and the release note, with the check group at the centre.
The red line: Claude drafts and checks, but you decide what the learner needs, you run through the course yourself before anyone else does, and a named person signs it off. Nothing invented is passed off as real, and no learner’s result is ever used to grade them.
Each skill holds one method or one artifact, with the inputs it needs, where it breaks, a fixed output template, a finish line and the next skill to run.
On facts, Claude flags and never fixes: it lists what it cannot trace to a source you approved, and the expert corrects it. The package and the tracking happen in your authoring tool and your LMS.
Download all 32 skills. One zip: ready-to-install skill zips, readable SKILL.md files, and the sources behind every method. Free, no signup. Download the zip · Browse on GitHub
The 32 skills you get
The set follows one course in eight stages: scope the course, get the content from experts, design it, write it, build it, check it, ship and improve, and your practice.
1 · Scope the course
1. Course Brief
Use when: "Can you just make it a course?" arrives with the format and the date already chosen
Output: The request restated as a problem, what is fixed and open, a "course, job aid or neither" line, a reply the requester can sign
2. Learner Analysis
Use when: The brief says "all staff" and nobody has asked who will sit in front of the screen
Output: Learners by role and task, device, place and real time, each line marked evidence or assumption, five questions for real learners
3. Action Mapping Workshop
Use when: The sponsor's list is everything people should know and nothing they should do
Output: A one-hour run of show, the map (goal, actions, causes, practice, minimum information), replies to "but they need to know", the cut list
4. Course Estimate and Timeline
Use when: The sponsor expects it next week "because AI writes it now"
Output: Effort by phase from your own past builds, three scope options for the date, review gates, the scope line the sponsor signs
2 · Get the content from experts
5. SME Interview Script
Use when: You asked the expert for answers and got a slide deck, or pages an AI wrote
Output: A pre-interview note asking for one real case, questions in three passes, a live notes sheet, the decision-point table, the follow-up message
6. Source Content Triage
Use when: The source is a 147-slide deck, two manuals that disagree, and a page nobody can date
Output: An inventory with dates, owners and conflicts, the practise, remember, look up, cut table, AI-written passages marked, a cut list the expert accepts
7. Review and Sign-Off Plan
Use when: Six people will review the course and nobody has said who breaks a tie
Output: Who reviews what at each gate, one named decider, rounds and dates, the change rule, the sign-off table
3 · Design it
8. Measurable Learning Objectives
Use when: The only objective on the page is "understand the policy"
Output: One objective per action with condition, verb and criterion, the level the job needs, the evidence each will need
9. Course Design Document
Use when: The sponsor sees finished screens and none of the decisions that made them
Output: Need, audience, strategy and practice per objective, format choices with reasons, the decisions log, the sign-off block
10. Course Curriculum Map
Use when: The course grew lesson by lesson and nothing connects
Output: Modules and lessons in order, objective, practice and check per lesson, the gaps, the learner's path
11. Assessment Blueprint
Use when: The quiz is written last, from whatever is on the slides
Output: Objectives by level, evidence type and number of items, where practice ends and assessment begins, what a result is used for
4 · Write it
12. Branching Scenario Script
Use when: Learners pass every screen and still freeze on the real call
Output: The setup, numbered decision nodes, options from real mistakes, consequence and feedback for each, a path map with every ending
13. E-Learning Storyboard Template
Use when: The build starts from the expert's slides and ends as narration over bullet points
Output: One row per screen with text, visual, narration, interaction, build note and alt text, coverage by objective, the cut list
14. Training Video Script
Use when: The script reads well on paper and sounds like a manual when said aloud
Output: A two-column script (heard, seen), a spoken-word rewrite, the shot list, caption and transcript text
15. Knowledge Check Questions
Use when: An AI can pass your quiz from a screenshot, so it proves nothing
Output: Items written to the blueprint as short decisions, distractors from real mistakes, feedback for every option, the retake rule
16. Job Aid Template
Use when: People need the answer in the middle of the task, and a module is the wrong place for it
Output: A checklist, decision table or one-page flow, the mistake next to its step, where it sits in the work, who updates it
5 · Build it
17. Interactive Practice Activity
Use when: You want to try a practice idea this afternoon, before anyone opens the authoring tool
Output: One practice activity as a Claude artifact to prototype and pilot, the spec behind it, a test log, a plain note on where it can live
18. Course Slides Build
Use when: Part of the course is delivered as slides, or your authoring tool starts from a deck
Output: A slide-by-slide deck from the storyboard, notes kept apart from slide text, the practice slides, an export check
19. Microlearning Series
Use when: Nobody will finish the long module, and chopping it into pieces is not the answer
Output: Short units of one action each, a retrieval prompt and a decision in each, the order and the gaps, what the series replaces
20. Live Session to E-Learning Conversion
Use when: "Put the workshop online" means a recording of slides
Output: What the room did, what replaces each part on screen, what must stay live, the new outline, what is lost and who accepts that
6 · Check it
21. AI Content Accuracy Check
Use when: The draft reads beautifully and nobody can say where its numbers came from
Output: Every factual claim traced to an approved source, untraced and conflicting items flagged for the expert, the expert's decision per flag
22. Does It Teach Review
Use when: The course looks finished and you suspect it teaches nothing
Output: Per objective: shown, tried, feedback that explains; screens that tell against screens that ask; fixes for the designer to accept or reject
23. Accessibility Check
Use when: Accessibility is found after the build, when every fix costs a rebuild
Output: A first pass marked pass, fail or cannot check here, the list a person must test with assistive technology, fix owners
24. Designer Run-Through
Use when: You built it, you reviewed it screen by screen, and you have never once taken it
Output: A protocol for taking the whole course yourself as a named kind of learner, an issues log with severity, your own verdict
25. SME Review Round
Use when: Comments arrive in four formats, two contradict each other, and one reopens the objectives
Output: The brief for one round, a comment sheet by lane, comments consolidated with conflicts side by side, a change log, the reply
26. Learner Pilot Test
Use when: The course is about to go to everyone and not one learner has tried it
Output: Who to invite by role, the "we are testing the course, not you" wording, a think-aloud script, findings by screen
7 · Ship and improve
27. Release Sign-Off Record
Use when: Launch is tomorrow and "it was reviewed" is the only evidence
Output: The result of each check with its date and who ran it, open issues with owners, the decision: release, fix first, or stop
28. Course Release Note
Use when: The course goes live with a file name and nothing else
Output: Purpose and audience, version, what changed in the learner's terms, known limits, owner and review date, the upload note
29. Course Data Review
Use when: The report says nearly everyone completed it, and that is all anyone reads
Output: Questions most people miss and the option that draws them, where people leave, comments in aggregate, a ranked fix list
8 · Your practice
30. Course Project Setup
Use when: Every chat starts from zero and Claude fills the gaps with plausible detail
Output: A setup sheet for one course in Claude: approved sources, standing instructions, what never goes in, a version naming rule
31. Learning Content Style Guide
Use when: Three courses sound like three different organisations
Output: Voice and reading level, terms, how feedback is worded, rules for examples, a before and after pair per rule
32. Portfolio Case Study
Use when: Your portfolio is screenshots, and the decisions behind them are what make you worth hiring
Output: The problem, your decisions and why, what you checked and changed, what Claude drafted and what you did, a confidentiality check
Where to start
| Your situation | Skill to run |
|---|---|
| “Can you just make it a course?” arrived with the format and the date already chosen? | Course Brief |
| The brief says “all staff” and nobody knows who will actually take it? | Learner Analysis |
| The sponsor expects it next week “because AI writes it now”? | Course Estimate and Timeline |
| You asked the expert for a real case and got a slide deck? | SME Interview Script |
| A pile of source material, with two manuals that disagree? | Source Content Triage |
| Six reviewers and nobody has said who decides? | Review and Sign-Off Plan |
| The only objective on the page is “understand the policy”? | Measurable Learning Objectives |
| Need the checks planned before the content is written? | Assessment Blueprint |
| Learners pass every screen and still freeze on the real call? | Branching Scenario Script |
| Need the course screen by screen before the build starts? | E-Learning Storyboard Template |
| Want to try a practice idea today, before anyone opens the authoring tool? | Interactive Practice Activity |
| The draft reads well and nobody can say where its numbers came from? | AI Content Accuracy Check |
| The course looks finished and you suspect it teaches nothing? | Does It Teach Review |
| You built it and have never once taken it yourself? | Designer Run-Through |
| About to go to everyone and not one learner has tried it? | Learner Pilot Test |
| Launch is tomorrow and “it was reviewed” is the only evidence? | Release Sign-Off Record |
The quality bar
Every skill in the pack holds the same standard:
- One method or one artifact, applied properly: each skill uses the real mechanics of its method (the gates, the passes, the grid, the questions asked at each step), not a generic “gather, analyse, recommend”
- It tells you where it breaks: every skill says when not to use it and which skill in the pack fits instead
- Same shape every time: a fixed output template, a short “done when” check, and a decision that a named person makes
- Flag, never fix, on facts: Claude lists what it cannot trace to a source you approved. The expert corrects it. Cases and characters are marked real or fictional
- Courses are checked, people are not: results are read by question and in aggregate to fix the course. Nothing grades, ranks, scores or profiles a learner, a reviewer or an expert
- Exact about Claude: skills name a Claude surface only where it helps, and say plainly that nothing here exports a standard e-learning package or reports to an LMS. Legal and accessibility-law questions end with “check with a qualified adviser”
- The red line: Claude drafts and checks, but you decide what the learner needs, you run through the course yourself before anyone else does, and a named person signs it off; nothing invented is passed off as real, and no learner’s result is ever used to grade them
Who made this
Polar Bear is a people ops consultancy for human-size teams (20 to 200 people). Built by ex-McKinsey founders with a dream to make AI work for People, not instead of them. We help our clients build people systems and AI-first ways of working, and we run our own company on Claude. This pack is the free, self-serve version of how we work.
This pack is for whoever is accountable for the course. If you run a learning team or build courses for clients and want it working this way, with AI carrying the brief, the storyboard, the scripts and the first pass of every check so people decide what the learner needs, take the course themselves and sign it off, that is what we build with clients.
I help instructional designers and L&D teams draft faster with Claude while people keep the judgement, the checks and the sign-off. Find 30 min together
Install in Claude Code
Two lines, and every skill in the pack loads at once.
/plugin marketplace add polar-bear-org/claude-skills /plugin install instructional-designers-pack@polar-bear-skills
Using Claude on the web instead? Download the zip and upload each skill from its install folder under Customize → Skills.
How to use it: pick the workflow that matches your week
Start with the course brief: it restates the request as a problem before anything is designed. After that, bring the one real thing in front of you.
Scope the request
A request arrived with the format and the date already chosen.
Course Brief → Learner Analysis → Action Mapping Workshop → Course Estimate and Timeline
Get the content from experts
The expert sent slides, and six people will review the course.
SME Interview Script → Source Content Triage → Review and Sign-Off Plan
Design before you build
The only objective is “understand the policy” and the quiz would be written last.
Measurable Learning Objectives → Course Design Document → Course Curriculum Map → Assessment Blueprint
Write the course
The blueprint is agreed and the screens are still blank.
Branching Scenario Script → E-Learning Storyboard Template → Training Video Script → Knowledge Check Questions → Job Aid Template
Prototype and build
You want to try a practice idea before anyone opens the authoring tool.
Interactive Practice Activity → Course Slides Build → Microlearning Series → Live Session to E-Learning Conversion
Check it before it ships
A draft exists and nobody has checked it or taken it.
AI Content Accuracy Check → Does It Teach Review → Accessibility Check → Designer Run-Through → SME Review Round → Learner Pilot Test
Ship and read the data
Launch is tomorrow and “it was reviewed” is the only evidence.
Release Sign-Off Record → Course Release Note → Course Data Review → Course Project Setup
Install it in five minutes
In Claude Code
/plugin marketplace add polar-bear-org/claude-skills /plugin install instructional-designers-pack@polar-bear-skills
In Claude.ai or Claude Desktop
- Download the pack and unzip it: its skills folder holds all 32 skills (https://meet-polar-bear.com/skills/instructional-designers-pack.zip). In Claude, click your name at the bottom left, then Settings.
- Under Customize, choose Skills. Click Add, then Upload a skill.
- Drag the skill's file into the box, or click to browse. You can add several at once.
- In the pack folder, open skills, then the skill you want, and select its SKILL.md (isd-course-brief is a good first one).
- Claude shows the skill's name and what it does. Click Upload.
- The skill opens on its own page. Its toggle should be on.
- Try it in a new chat: type / and the skill's name, then say what you need.
Then try one
Run isd-course-brief on the email asking for the course: [paste it here]Run isd-ai-accuracy-check on this storyboard against my approved sources: [paste the storyboard and the sources here]Run isd-does-it-teach-review on this draft against these objectives: [paste them here]
Questions people ask
What is in the Claude for Instructional Designers pack?
32 Claude skills that follow one course in eight stages: scope it, get the content from experts, design it, write it, build it, check it, ship and improve, and your practice. Each skill holds one method or one artifact.
How do I install these Claude skills?
In Claude Code, two lines install all 32 at once. In Claude.ai or Claude Desktop, download the zip, open Settings, Skills, Add, and upload the skills you want. Either way it takes about five minutes.
Do I need Claude Code to use them?
No. The skills work in Claude.ai, Claude Desktop and Claude Code. In the app you upload them yourself; in Claude Code one command installs the whole pack.
Do the skills work with ChatGPT or other assistants?
They are built as Claude Skills, and installing them is a Claude feature. Each one is a plain Markdown file, so you can read it and reuse the method elsewhere, but it will not load and trigger by itself outside Claude.
Does Claude sign off the course or grade learners?
No. Claude drafts and checks, but you decide what the learner needs, you run through the course yourself before anyone else does, and a named person signs it off. Nothing invented is passed off as real, and no learner’s result is ever used to grade them.
Is the pack free?
Yes. It is free to use inside your company and in client work, and not for resale. There is no signup: the download and the install command are on this page.