assess-slts

name	assess-slts
description	Evaluate SLT quality across 5 dimensions with rewrite suggestions.
license	MIT
metadata	{"author":"Andamio","version":"1.0.0"}

Skill: Assess Student Learning Targets

Description

Evaluates the quality of a list of Student Learning Targets (SLTs) against established education research and Andamio protocol conventions. Provides per-SLT assessment and actionable suggestions for improvement.

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 assessing SLTs, read the knowledge base for patterns that should influence your assessment. 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/quality-issues.yaml

   • Flag SLTs matching known problematic patterns
   • Include "seen in [n] courses" when citing issues

2. Read knowledge/slt-patterns/successful-rewrites.yaml

   • When suggesting fixes, prefer proven rewrites
   • Include "this rewrite worked in [course]" when applicable

3. Surface knowledge in assessment (if relevant patterns exist):

   **Known Issue:** This pattern (task-focused verb) was flagged in 3 previous courses.
   **Proven Rewrite:** "I can complete the setup" → "I can configure my environment to run X" (from andamio-for-contributors)
   

---

The user will provide a markdown file containing a list of SLTs. Read the file, then assess each SLT against the criteria below. Always read knowledge/research/slt-research-report.md for full research context before running the assessment.

Assessment Criteria

Evaluate each SLT on the following dimensions, rating each as Strong, Acceptable, or Needs Work:

1. Student-Facing Language & Approachability

• Is it written from the learner's perspective?
• Does it use "I can..." or "Learner can..." format?
• Is the language accessible and free of unnecessary jargon?
• Would a motivated beginner read this SLT and feel invited into the topic rather than intimidated by it?
• Tool names, library names, and framework names are fine — they orient the learner. But implementation details (specific functions, flags, protocol specifics) belong in the lesson, not the SLT.
• Strong: Clear "I can..." statement that any learner could read and immediately understand what they'll be able to do. Names the relevant tools/concepts without burying the reader in technical detail.
• Acceptable: Learner-facing but leans on jargon or implementation specifics that narrow the audience unnecessarily.
• Needs Work: Written as a teacher objective, uses third person, or reads like a spec rather than a learning invitation.

2. Focus on Learning, Not Tasks

• Does it describe what the learner will learn or be able to do, rather than what activity they will perform?
• Strong: Clearly states a capability or understanding the learner will gain.
• Acceptable: Implies learning but is phrased partly as an activity.
• Needs Work: Describes a task or activity rather than a learning outcome (e.g., "I can complete the worksheet" or "I can read chapter 5").

3. Specificity and Measurability

• Is the SLT narrow enough that mastery can be assessed?
• Could a teacher (or peer) determine whether this target has been met based on evidence?
• Strong: Clear, observable, and assessable. You could design an assignment around it.
• Acceptable: Mostly assessable but somewhat broad or ambiguous.
• Needs Work: Too vague to assess (e.g., "I can understand blockchain") or too broad to fit a single module.

4. Cognitive Level (Bloom's Taxonomy)

• Identify the Bloom's level implied by the verb: Remember, Understand, Apply, Analyze, Evaluate, or Create.
• Is the cognitive level appropriate for the apparent subject matter and context?
• Flag SLTs that cluster entirely at low levels (Remember/Understand) when higher-order thinking would be expected.
• Flag SLTs that use vague verbs like "understand" or "know" — these are difficult to measure. Suggest more precise alternatives.

5. Standalone Clarity

• Can the SLT be understood without needing to read surrounding course material?
• Does it make sense as an independent unit of learning?
• Strong: Self-contained and clear.
• Acceptable: Understandable but references implicit context.
• Needs Work: Requires significant external context to interpret.

Output Format

Respond with a markdown document structured as follows:

# SLT Quality Assessment

## Summary

- **Total SLTs reviewed**: [number]
- **Overall quality**: [brief 1-2 sentence summary]
- **Distribution**: [count] Strong / [count] Acceptable / [count] Needs Work

## Per-SLT Assessment

### SLT 1: "[full text of SLT]"

| Criterion | Rating | Notes |
|-----------|--------|-------|
| Student-Facing Language | [rating] | [brief note] |
| Focus on Learning | [rating] | [brief note] |
| Specificity & Measurability | [rating] | [brief note] |
| Cognitive Level | [Bloom's level] | [brief note] |
| Standalone Clarity | [rating] | [brief note] |

**Overall**: [Strong / Acceptable / Needs Work]

**Suggestion**: [If not Strong, provide a concrete rewrite or improvement. If Strong, say "No changes needed."]

---

[Repeat for each SLT]

## Set-Level Observations

[After all individual assessments, note any patterns across the full set:]
- Bloom's level distribution — are targets clustered at one level?
- Gaps — are there obvious areas of learning not covered?
- Redundancy — do any SLTs overlap significantly?
- Prerequisite logic — does the ordering suggest a reasonable learning progression?

Guidelines

• Be direct and specific. Vague feedback like "could be better" is not useful.
• When suggesting a rewrite, provide the full rewritten SLT text.
• Preserve the learner's domain and intent — don't change what they're learning, improve how it's expressed.
• If an SLT is Strong on all dimensions, say so briefly and move on. Spend more space on SLTs that need work.
• The overall rating for an SLT is determined by its weakest dimension: one "Needs Work" rating makes the overall "Needs Work."

Tone & Technicality

• SLTs should be approachable gateways, not technical specifications. They should make a learner curious about what's ahead, not anxious about what they don't yet know.
• Name tools and frameworks (e.g., "Cobra", "Fiber", "Go") — these give learners real-world orientation. But leave implementation details (specific functions, flags, API patterns, protocol internals) to the lesson content.
• A good test: if someone who has never touched the subject reads the SLT, they should understand the shape of what they'll learn, even if they can't do it yet.
• When rewriting an overly technical SLT, aim for language that is specific enough to assess but plain enough to invite. Err on the side of approachability — the lesson itself is where depth lives.