ielts-listening-review

Diagnose IELTS listening errors and generate validated DOCX error notebooks with replay drills.

Updated Apr 13, 2026
One-click install
npx skills add https://github.com/xiaoshan1234/ai-skill --skill ielts-listening-review-xiaoshan1234
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ielts-listening-review
Source: https://github.com/xiaoshan1234/ai-skill/tree/main/role/english-teacher/skills/ielts-listening-review
Command: npx skills add https://github.com/xiaoshan1234/ai-skill --skill ielts-listening-review-xiaoshan1234

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? IELTS learners often finish a listening or dictation practice without understanding why they missed answers, and generic advice like "listen more" does not fix specific weaknesses. This Skill analyzes submitted answers, transcripts, and timestamps to classify each error, design targeted replay drills, and produce a validated Listening Error Notebook DOCX. ## Core Features & Use Cases - Error Classification: Categorizes each missed item into concrete causes such as weak forms, connected speech, spelling, singular/plural, numbers, distractors, paraphrase, or grammar fit, mapped to a canonical skill taxonomy. - Intensive Listening Drills: Builds 1-2 minute micro-exercises per question (shadowing, spelling, singular/plural checks, distractor identification) with immediate retesting after browser-based replay. - Validated DOCX Notebook: Generates a Listening Error Notebook DOCX with an error map, per-item analysis, review plan, and vocabulary, then validates structure, fonts, and plan content before delivery. - Learning Progress Tracking: Records objective and retrieval evidence into a local SQLite mirror synced with the IELTS Buddy cloud event log, driving spaced review scheduling. - Use Case: After submitting an IELTS Buddy listening session, ask the Skill to review your mistakes; it reads the session via MCP, explains why each answer was wrong, assigns replay targets, and saves a DOCX notebook to your Desktop. ## Quick Start Review my submitted IELTS listening session, explain why I missed each answer, and create a Listening Error Notebook DOCX with replay drills.

Frequently Asked Questions about ielts-listening-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review my IELTS listening mistakes after a practice session?▼

Submit your listening session in the IELTS Buddy browser first, then ask for a review. The Skill reads the submitted session via MCP, classifies each error by cause, assigns a replay target and micro-drill per question, and can generate a DOCX error notebook.

What types of listening errors can be diagnosed?▼

Errors are classified into sound discrimination, weak forms or connected speech, spelling, singular/plural, numbers and dates, word boundaries, distractors, paraphrase, grammar fit, and attention loss. Each cause maps to a canonical code in the bundled skill taxonomy.

Does this Skill play audio or run timed listening practice?▼

No. Audio playback, timing, and answer entry are browser-owned and happen in the IELTS Buddy web interface. The Skill works after submission, analyzing evidence and designing replay drills rather than recreating the player.

Can it work without the IELTS Buddy MCP connection?▼

Yes. Without MCP, it runs local workflows using transcripts, answers, and notes you paste directly. MCP is only needed for reading submitted sessions, cloud progress sync, and vocabulary saving.

How is the Listening Error Notebook DOCX validated?▼

A bundled validation script checks the DOCX for required sections, at least one table, Times New Roman fonts, and that every plan item's correct answer, error type, replay target, and micro-drill appears in the document. Failed validation triggers regeneration.

How does spaced review scheduling work for listening errors?▼

Objective and retrieval results are recorded as learning events in a local SQLite mirror synced to the cloud. Correct answers schedule reviews at 1, 3, 7, 14, 30, 60, and 90 day intervals, while incorrect results become due immediately.