Draft Slts

Skill: Draft Student Learning Targets

Description

Generates well-formed Student Learning Targets (SLTs) for a course or module based on topic, audience, and learning goals. Follows education research best practices and Andamio protocol conventions.

Instructions

Path Resolution

Resolve file paths based on your execution context:

• Plugin context (${CLAUDE_PLUGIN_ROOT} is set): Read knowledge from ${CLAUDE_PLUGIN_DATA}/knowledge/ (user data), falling back to ${CLAUDE_PLUGIN_ROOT}/knowledge/ (seed data). Read research from ${CLAUDE_PLUGIN_ROOT}/knowledge/research/.
• Clone/symlink context (default): Read knowledge from knowledge/ relative to the project root (research is at knowledge/research/).

Pre-Execution Knowledge Check

Before drafting SLTs, read the knowledge base for patterns that should influence your work. If any knowledge file does not exist, skip it and proceed without prior patterns — note "No prior data available" in output.

1. Read knowledge/slt-patterns/verb-bank.yaml

• Note effective verbs by Bloom's level
• Prefer verbs with high success_count

1. Read knowledge/slt-patterns/quality-issues.yaml

• Note problematic patterns to avoid
• Especially avoid patterns with high frequency

1. Surface knowledge to user (if relevant patterns exist):

```markdown ### Tips from Previous Courses

Effective verbs at this level: compare, distinguish, implement (12 successes) Patterns to avoid: "understand X" (flagged 8 times), task-focused phrasing ```

---

The user will describe a course or module they want to create. Gather the following information (ask if not provided):

1. Topic: What is the course/module about?
2. Audience: Who are the learners? What's their background?
3. Learning goals: What should learners be able to do after completing this?
4. Module count: How many modules? (Default: 3-4)
5. SLTs per module: How many SLTs per module? (Default: 2-4)

Always read knowledge/research/slt-research-report.md for full research context before drafting.

The SLT Formula

Every SLT must follow this structure:

I can + [measurable verb] + [specific outcome] + [observable evidence/context]

Examples:

• Weak: "I can understand blockchain" (unmeasurable, vague)
• Better: "I can explain what a blockchain is" (measurable, but no evidence)
• Strong: "I can compare a Portable Access Token to a traditional platform account by identifying at least two differences in ownership, portability, and revocability" (measurable verb, specific outcome, observable evidence)

Verb Selection (Bloom's Taxonomy)

Use verbs that match the cognitive level appropriate for the content:

| Level | Verbs | When to Use | |-------|-------|-------------| | Remember | list, name, identify, recall | Foundational facts, terminology | | Understand | explain, describe, summarize, compare | Conceptual understanding | | Apply | use, demonstrate, implement, execute | Procedural skills | | Analyze | distinguish, examine, compare, contrast | Breaking down complex topics | | Evaluate | judge, justify, critique, assess | Making judgments | | Create | design, construct, develop, produce | Building something new |

Progression principle: Early modules should include more Remember/Understand/Apply. Later modules can include Analyze/Evaluate/Create.

Verbs to Avoid

Never use these unmeasurable verbs:

• understand, know, learn, appreciate, be aware of, become familiar with

These cannot be assessed. Replace with specific, observable actions.

Output Format

Generate SLTs in this exact format:

# SLTs: [Course Name]

## Module 1: [Module Title]

- 1.1: I can [verb] [specific outcome] [evidence/context].
- 1.2: I can [verb] [specific outcome] [evidence/context].
- 1.3: I can [verb] [specific outcome] [evidence/context].

## Module 2: [Module Title]

- 2.1: I can [verb] [specific outcome] [evidence/context].
- 2.2: I can [verb] [specific outcome] [evidence/context].

## Module 3: [Module Title]

- 3.1: I can [verb] [specific outcome] [evidence/context].
- 3.2: I can [verb] [specific outcome] [evidence/context].
- 3.3: I can [verb] [specific outcome] [evidence/context].

Guidelines

1. One capability per SLT. If an SLT has "and" connecting two different skills, split it.

1. Make it assessable. Ask: "Could a teacher design an assignment to verify this?" If not, rewrite.

1. Write for the learner. SLTs should make learners curious, not anxious. Name tools and frameworks, but leave implementation details to the lessons.

1. Build progression. Module 1 SLTs should be achievable early. Later modules can build on earlier ones.

1. 2-4 SLTs per module is ideal. More than 5 per module suggests the module should be split.

1. Consider the assignment. Each module needs one assignment that can assess all its SLTs. If the SLTs don't fit a coherent assignment, reorganize.

1. Mark optional modules. If a module is optional/advanced, note it in the title: "Module 4: [Title] (Optional)"

Quality Checklist

Before outputting, verify each SLT:

• [ ] Starts with "I can"
• [ ] Uses a measurable verb (not understand/know/learn)
• [ ] Has a specific outcome (not vague)
• [ ] Includes observable evidence or context
• [ ] Can be assessed through an assignment
• [ ] Is appropriate for the stated audience
• [ ] Fits the cognitive progression (Bloom's levels build across modules)

After Drafting

Suggest the user:

1. Review and refine the SLTs with their team
2. Run /assess-slts to get quality feedback
3. Create 00-course.md with lesson types and assignments based on the finalized SLTs

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: Andamio-Platform
• Source: Andamio-Platform/coach
• License: MIT

Install and usage instructions live in the source repository linked above.