review

Review — Learning Progress Dashboard

ALWAYS use the AskUserQuestion tool when asking the user questions, in any context. If you have too many questions for the tool, split them up into multiple calls.

Overview

Show the user their learning progress across all topics or for a specific topic. Display quiz scores, module completion, weak areas, score trends, and actionable next steps.

Data Schemas

The review skill reads data written by the learn and quiz skills. Here are the formats you'll encounter:

index.json (~/.claude/learning/index.json) — master registry:

{
  "topics": {
    "topic-slug": {
      "displayName": "Topic Name",
      "planFile": "plans/topic-slug-2026-03-01.json",
      "progressFile": "progress/topic-slug.json",
      "created": "2026-03-01",
      "lastActivity": "2026-03-01",
      "level": "beginner",
      "modulesCompleted": 0,
      "modulesTotal": 5,
      "quizzesTaken": 1,
      "overallScore": 0.8
    }
  }
}

Plan files (~/.claude/learning/plans/<slug>-<date>.json) — contain modules[] with id, title, objectives, keyConcepts, estimatedTime, resources[].

Progress files (~/.claude/learning/progress/<slug>.json) — contain quizzes[] (each with date, module, score, total, questions[]), weakAreas[], strongAreas[], overallScore (percentage 0-100, not fraction).

Progress files now also contain spacedRepetition — a map of concept labels to SM-2 scheduling data (easeFactor, intervalDays, nextReview, repetitions). Concepts with nextReview at or before today are due for review.

Process

Step 1: Load Data

Read ~/.claude/learning/index.json.

If file doesn't exist or is empty:

• Tell the user: "No learning data found yet. Use /learn to start your first topic!"
• Stop here.

If user specified a topic (e.g., /review kubernetes):

• Load that topic's plan and progress files
• Show detailed single-topic view

If no topic specified:

• Load all topics from index
• Show overview of all topics

Step 2: Display Progress

All-topics overview format:

── Learning Progress ─────────────────────────────

[Topic Name] (started [date])
  Module 1: [Title]    [progress bar] [score]%  ([N] quizzes)
  Module 2: [Title]    [progress bar] not started
  ...
  Weak areas: [list]
  Next step: [actionable suggestion]

[Topic Name 2] ...

── Due for Review ────────────────────────────────
  [Topic]: [N] concepts due ([list top 3 concept names])
  [Topic 2]: [N] concepts due ([list top 3])

  Run /quiz [topic] to review these concepts.
──────────────────────────────────────────────────

── Summary ───────────────────────────────────────
  Active topics: [N]
  Total quizzes taken: [N]
  Average score: [N]%
  Suggested focus: [topic — reason]
──────────────────────────────────────────────────

Due for review section:

If any concepts have nextReview at or before today across any topic, show the "Due for Review" block between the per-topic details and the Summary (as shown above). If no concepts are due, omit this block entirely.

Progress bar rendering:

• Use Unicode block characters: █ for filled, ░ for empty
• 10 characters wide
• Based on quiz scores for that module (no quizzes = all empty)

Single-topic detailed format:

• Show all modules with per-module quiz history
• Show score trend if multiple quizzes (e.g., "Module 1: 60% → 75% → 90%")
• List all weak and strong areas
• Show total time estimated vs modules completed
• Suggest specific module to focus on next

Step 3: Recommend Next Action

Always end with one clear, actionable suggestion:

• If concepts are due for review (nextReview <= today): "You have [N] concepts due for review in [topic]. Run /quiz [topic] to keep them fresh — spaced repetition works best when you review on schedule."
• If a topic has weak areas: "Retake the quiz on [topic] focusing on [weak area]. Use /quiz [topic]"
• If a module is unstarted: "Start Module [N]: [title]. Check /resources [topic] for study materials"
• If all modules are strong: "Great work on [topic]! Consider exploring a related topic or increasing the difficulty"
• If multiple topics exist: prioritize the one with lowest score or most staleness