Open Source · scibly-dev/skills
AI Skills for Instructional Designers and Course Creators
Nine specialized agent skills for L&D teams. From needs analysis to storyboard, for every phase of course design.
npx skills add scibly-dev/skills --all
needs-analysis
Analysis
What needs to be learned?
Turns SME notes, transcripts, and job descriptions into a structured skill gap analysis. Raw input becomes a learning brief you can build from immediately.
When to use
When it's unclear what a course should actually cover. Most useful after SME interviews or before scoping a new training project.
didactic-reduction
Analysis
Cut everything that doesn't need to be taught.
Filters any topic down to its didactically essential core. Uses three filters from educational science to decide what learners must understand and what can be left out without losing integrity.
When to use
When an SME has delivered 200 slides and it's unclear what belongs in the course. Or when training content is too broad for the time available.
learning-objectives
Course design
From vague to measurable.
Turns weak objectives like "employees understand the process" into precise, testable behavior statements using Bloom's Taxonomy and the ABCD model.
When to use
Directly after needs analysis or before structuring a course. Without clear objectives, assessments test the wrong things and microlearning covers the wrong topics.
microlearning-design
Course design
Structure for 3–10 minute learning moments.
Designs focused microlearning sessions: which scenes, what sequence, which practice moment. The skill structures the learning unit, not just the content.
When to use
When a course needs to be split into short, mobile-friendly units. Especially useful for onboarding, performance support, and just-in-time training.
assessment-design
Course design
Quiz questions that actually test understanding.
Creates questions that go beyond simple recognition. Uses Bloom's Taxonomy to generate application-level items with working distractors and feedback that continues teaching.
When to use
When AI-generated questions are too easy, too obvious, or test only recall. Or when a course needs a valid completion record.
scenario-generation
Content
Realistic workplace scenarios at scale.
Generates realistic scenarios, case studies, and branching situations. With the right brief, AI produces ten scenarios in the time it takes to write one by hand.
When to use
When a skill needs many practice situations, or when existing scenarios feel too generic or unrealistic.
content-editing
Content
AI writes the draft. You make it good.
Guides the review of AI-generated content: what to look for, how to fix it fast. Cuts development time without sacrificing quality.
When to use
When there's AI output that feels off but it's hard to say why. Or when a team wants to use AI as a first-draft machine without losing editorial control.
instructional-prompt-engineering
Content
Better prompts, better course content.
Shows how to structure AI prompts for learning content using a 4-ingredient framework that delivers pedagogically sound, audience-appropriate output instead of generic text.
When to use
When AI output feels generic or off-target for the audience. Or when someone wants to learn how to brief AI specifically for instructional design tasks.
storyboard
Production
What do we produce, and where does it make sense?
Turns a course outline into a media plan: which scene gets a video, animation, interactive scenario, or plain text, with didactic reasoning for each decision.
When to use
When a course structure exists and the question arises: what do we actually produce? Most useful before video or animation production begins.