메뉴
Read a newly uploaded file from the inbox, classify it, move it into the right subject/topic folder, and write a metadata JSON.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Build a designed, self-contained HTML academic map of the child's subjects, topics, and materials, plus real coaching notes for the parent on how to teach them better.
Build a designed, self-contained HTML progress report for the child — how she's actually doing, and what to work on next — surfaced in the Progress tab for both parent and child.
Create clear, child-ready study material (notes, worked examples, a revision sheet) for a topic, matched to the child's level and their own uploaded materials, packaged as one activity.
Create a practice test for the child from their materials and progress — child-facing questions with a separate parent-only answer key, packaged as one activity.
Extract genuine interest signals from the child's own conversations and keep memory/interests.md current — a Pulse-only skill, not triggered by a direct parent request.
Extract durable things the parent has said in chat and keep memory/preferences.md current — a Pulse-only skill, not triggered by a direct parent request.
| name | process-file |
| description | Read a newly uploaded file from the inbox, classify it, move it into the right subject/topic folder, and write a metadata JSON. |
Whenever there are files in inbox/, process each one before doing anything else:
Read it. First file "inbox/<name>" to see what it really is, then extract its actual content by following skills/read-file/SKILL.md — the canonical guide for reading any format (PDF via pypdf; scanned/complex/OCR via liteparse; Word/PowerPoint via pandoc; Excel via openpyxl; images via read_image; archives, audio/video, etc.). Only record content you actually extracted — never invent what you couldn't read.
Classify it from the content (or the parent's description):
subject — e.g. Mathematics, Science, English.topic — the specific topic, e.g. "quadratic equations".type — one of: notes, worksheet, textbook-page, homework, test, image, other.Be interactive — ask when in doubt. Check materials/ for subjects/topics that already exist. If the file's content doesn't clearly match any of them, looks ambiguous, or you cannot confidently classify it, do NOT guess: leave the file in inbox/, tell the parent what you see, and ask them which subject/topic it belongs to. Move and record it only once you are confident (or the parent has told you). It is always better to ask than to mis-file.
Move it into the proper folder, keeping the original filename:
mkdir -p "materials/<subject>/<topic>"
mv "inbox/<filename>" "materials/<subject>/<topic>/<filename>"
Write metadata next to it at materials/<subject>/<topic>/<filename>.meta.json — include the FULL text you extracted in step 1 as extracted_text, not just a summary. This is what makes step 1's read-image/OCR/liteparse work a one-time cost: any skill that needs this material later (create-test, create-study-material, create-academic-map, or you in a future chat) reads extracted_text straight from this file instead of re-running vision/OCR on the original again:
{
"original_name": "<original filename>",
"stored_path": "materials/<subject>/<topic>/<filename>",
"subject": "<subject>",
"topic": "<topic>",
"type": "<type>",
"summary": "<1-2 sentence summary of what the file contains>",
"key_concepts": ["<concept>", "<concept>"],
"extracted_text": "<the full text/transcription you extracted in step 1 — verbatim, not a paraphrase>",
"source": "parent-upload",
"processed_at": "<run: date -u +%Y-%m-%dT%H:%M:%SZ>"
}
Tell the parent, in plain words, what you filed and where, and confirm the subject/topic you chose so they can correct you if it is wrong.