Listening Error Review

Listening Error Review

Core Rule

Listening playback and dictation input are browser-owned. Local Agent value starts after the attempt: classify errors, build replay targets, and schedule drills.

Inputs

Use any available combination:

• browser result summary;
• transcript or dictation text;
• user answer and correct answer;
• audio timestamp or section number;
• question type and answer key;
• user-pasted notes from the web session.

If the user only asks to do dictation, provide the listening-dictation route from references/web-workspace.md.

Error Taxonomy

Classify each miss into the main cause:

• sound discrimination;
• weak form or connected speech;
• spelling;
• singular/plural;
• number, date, or unit;
• word boundary;
• distractor;
• paraphrase;
• grammar fit;
• attention/location loss.

Avoid generic “listen more” feedback.

Workflow

1. List missed items only.
2. For each item, show correct answer, user answer, error type, and one replay target.
3. Create a 1-2 minute micro-drill:

• shadow the phrase;
• spell the answer;
• distinguish singular/plural;
• predict answer grammar before replay;
• identify distractor phrase.

1. Save high-value phrases to vocabulary only when the user asks or when the phrase is clearly reusable.
2. Record learning events after grading.

Default Deliverable

Deliver a validated Listening Error Notebook .docx by default when there are 2+ missed items, a full dictation review, or a section review. For one quick miss, concise chat is acceptable.

1. Build an error-notebook plan from transcript/result data.
2. Generate the DOCX with scripts/create_listening_error_notebook_docx.py.
3. Validate it with scripts/validate_listening_error_notebook_docx.py.
4. Return the absolute path to the final DOCX.
5. If validation fails, fix the plan and rerun generation plus validation.

Bundled Resources

• scripts/create_listening_error_notebook_docx.py: create a Listening Error Notebook DOCX with error map, item analysis, replay plan, micro-drills, and vocabulary.
• scripts/validate_listening_error_notebook_docx.py: verify required sections, table, Times New Roman, and plan content.

JSON Error Notebook Plan

{
  "source_title": "IELTS Listening Section 2",
  "section_title": "Community centre booking",
  "summary": "The main issue was plural endings and distractors after correction phrases.",
  "items": [
    {
      "question_number": "12",
      "user_answer": "ticket",
      "correct_answer": "tickets",
      "error_type": "singular/plural",
      "transcript": "You will need to bring your tickets to the entrance.",
      "cause": "The final /s/ was weak but grammatically required by bring your ...",
      "replay_target": "bring your tickets",
      "micro_drill": "Replay the phrase twice and mark whether the answer must be singular or plural before listening again."
    }
  ],
  "review_plan": ["Replay only the target phrases, not the whole audio."],
  "vocabulary": [
    {"phrase": "entrance", "meaning": "入口", "source_sentence": "bring your tickets to the entrance"}
  ]
}

python scripts/create_listening_error_notebook_docx.py listening_error_plan.json
python scripts/validate_listening_error_notebook_docx.py ~/Desktop/IELTS_Listening_Error_Notebook_YYYYMMDD_HHMM.docx --plan-json listening_error_plan.json

Source & license

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

• Author: Jobo16
• Source: Jobo16/ielts-all-in-one-skills
• License: MIT
• Homepage: https://ieltsbuddy.igocn.cn

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