← スキル一覧に戻る

course-report-analyzer
by TobiasBak
⭐ 0🍴 0📅 2026年1月17日
SKILL.md
name: course-report-analyzer description: | Use this skill when the user wants to analyze a student report against course lecture materials. Triggers when user asks to: "analyze my report against the course", "check course coverage", "compare report to lectures", "suggest report improvements based on course material", or "create lecture summaries and analyze report". version: 1.0.0
Course Report Analyzer
Analyzes student project reports against course lecture materials to identify coverage gaps and suggest improvements.
Workflow Overview
1. Discover Files → 2. Create Summary Dir → 3. Parallel Agents → 4. Read All → 5. Generate Suggestions
Step 1: Discover and Validate Files
Find all required files using Glob:
Lectures: **/*lecture*.pdf OR **/*lesson*.pdf OR **/lectures/**/*.pdf
Report: **/*report*.pdf
Requirements: **/*requirement*.txt OR **/*requirement*.md
Step 2: Create Summary Directory
mkdir summary
Step 3: Launch Parallel Background Agents
CRITICAL: Launch ALL agents in a SINGLE message for true parallelism.
For each lecture, use the Task tool with these parameters:
| Parameter | Value |
|---|---|
| subagent_type | general-purpose |
| run_in_background | true |
| description | Write lesson_XX summary |
Agent Prompt Template:
Read the lecture file(s) at [LECTURE_PATH] and write a concise summary to [SUMMARY_DIR]/lesson_XX_[topic].md
The summary MUST include these sections:
## Key Concepts
## Main Topics and Techniques
## Practical Applications
## Important [Engine/Framework] Features
If the lecture is split into multiple parts (part1.pdf, part2.pdf), read ALL parts before writing.
Keep it focused and practical for a [COURSE_TYPE] course.
Example - launching 10 agents in parallel:
<function_calls>
<invoke name="Task">
<parameter name="description">Write lesson_02 summary</parameter>
<parameter name="prompt">Read lectures/Lesson_02*.pdf, write to summary/lesson_02_blueprints.md...</parameter>
<parameter name="subagent_type">general-purpose</parameter>
<parameter name="run_in_background">true</parameter>
</invoke>
<invoke name="Task">
<parameter name="description">Write lesson_03 summary</parameter>
<parameter name="prompt">Read lectures/Lesson_03*.pdf, write to summary/lesson_03_cpp.md...</parameter>
<parameter name="subagent_type">general-purpose</parameter>
<parameter name="run_in_background">true</parameter>
</invoke>
<!-- ... repeat for all lectures ... -->
</function_calls>
Step 4: Wait and Collect Results
While agents run, you can work on other tasks. Check progress with:
<invoke name="TaskOutput">
<parameter name="task_id">[AGENT_ID]</parameter>
<parameter name="block">false</parameter>
</invoke>
When ready to wait for completion:
<invoke name="TaskOutput">
<parameter name="task_id">[AGENT_ID]</parameter>
<parameter name="block">true</parameter>
<parameter name="timeout">120000</parameter>
</invoke>
Step 5: Read All Content
After all agents complete, read in parallel:
- All lecture summaries - Use Read tool for each
summary/*.md - The report PDF - Read tool handles PDFs directly
- Requirements files - Both project and report requirements
Step 6: Generate report_suggestions.md
Create a comprehensive suggestions file with this structure:
# Report Improvement Suggestions
## Executive Summary
[3-4 bullet points on main improvement areas]
## Course Coverage Analysis
### Strong Coverage (Keep/Minor Enhancements)
| Lesson | Topic | Report Section | Notes |
### Partial Coverage (Should Expand)
| Lesson | Topic | What's Missing |
### Not Covered (Should Add)
| Lesson | Topic | Relevance to Project |
## Section-by-Section Suggestions
### [Section Name] (Page X)
**Current State**: [Assessment]
**Suggestions**:
- [Specific suggestion with example text]
### NEW SECTION: [Suggested Addition]
[Proposed content]
## Quick Wins - Easy Additions
| Location | Addition |
## Missing Topics to Address
### [Topic from course]
[Why relevant, suggested text]
## Terminology Reference
| Generic Term | Course-Specific Term |
## Summary
[Priority improvements list]
File Naming Conventions
| Type | Pattern | Example |
|---|---|---|
| Lecture Summary | lesson_XX_[topic].md | lesson_02_blueprints.md |
| Suggestions | report_suggestions.md | - |
| Summary Dir | summary/ | - |
Tips for Quality Results
- Agents should read ALL parts of split lectures before summarizing
- Summaries should be practical - focus on what's applicable to projects
- Cross-reference carefully - map report sections to specific lessons
- Include example text - don't just say "add X", show what to add
- Use course terminology - identify generic vs course-specific terms
- Be specific about pages/sections - reference exact locations in report
スコア
総合スコア
50/100
リポジトリの品質指標に基づく評価
✓SKILL.md
SKILL.mdファイルが含まれている
+20
○LICENSE
ライセンスが設定されている
0/10
○説明文
100文字以上の説明がある
0/10
○人気
GitHub Stars 100以上
0/15
○最近の活動
3ヶ月以内に更新がある
0/10
○フォーク
10回以上フォークされている
0/5
✓Issue管理
オープンIssueが50未満
+5
✓言語
プログラミング言語が設定されている
+5
○タグ
1つ以上のタグが設定されている
0/5
レビュー
💬
レビュー機能は近日公開予定です