ielts-speaking-coach
Simulate, score, and give feedback on IELTS Speaking tests (Parts 1, 2, and 3). Use this skill when the user wants to practise IELTS speaking, get sample answers, receive band-score feedback on a response they wrote or transcribed, prepare for the cue card (Part 2 long turn), or understand what examiners look for. Also use when the user asks about how to improve fluency, build vocabulary for speaking, reduce hesitation, or handle Part 3 abstract questions. Covers all four speaking criteria: Fluency and Coherence, Lexical Resource, Grammatical Range and Accuracy, Pronunciation.
How do I install this agent skill?
npx skills add https://github.com/jurgendn/agent-skills --skill ielts-speaking-coachIs this agent skill safe to install?
- Gen Agent Trust Hubpass
The skill is an IELTS Speaking test preparation tool designed to simulate interviews, provide band scores based on official criteria, and offer feedback on fluency, vocabulary, and grammar. No security risks, malicious patterns, or dangerous operations were detected.
- Socketpass
No alerts
- Snykpass
Risk: LOW · No issues
What does this agent skill do?
IELTS Speaking Coach
The IELTS Speaking test is a 11–14 minute face-to-face interview in three parts. It is assessed live or via recording by a trained examiner on four equally-weighted criteria. Unlike Writing, there is no minimum word count — but depth, naturalness, and range are rewarded.
Use this when
- The user wants to practise any part of the Speaking test (mock questions, sample answers, role-play).
- The user provides a transcript or written version of a spoken response and wants scoring feedback.
- The user asks how to improve fluency, extend answers, or use more natural language.
- The user wants to understand the scoring criteria or prepare for a specific topic.
Do not use this when
- The user is working on Writing Task 1 or Task 2. Use the appropriate writing skill.
- The user is preparing a formal presentation or research talk — use a presentation-specific skill.
Test Structure
| Part | Duration | Format |
|---|---|---|
| Part 1 — Introduction & Interview | 4–5 min | Familiar topics: home, work/study, hobbies, daily routine. Short answers expected (2–4 sentences). |
| Part 2 — Individual Long Turn | 3–4 min | Candidate speaks for 1–2 minutes on a cue card topic. 1 minute preparation time. |
| Part 3 — Two-way Discussion | 4–5 min | Abstract discussion linked to Part 2 topic. Extended, analytical answers expected. |
Band Descriptors (Speaking)
All four criteria are weighted equally. The overall Speaking band is the average, rounded to the nearest 0.5.
Full band-by-band reference (bands 5–9, all four Speaking criteria) with links to the official IELTS descriptors:
references/ielts-band-descriptors.md. Consult it before assigning a band so scoring stays anchored to the public criteria, not impression.
Fluency and Coherence (FC)
| Band | Descriptor |
|---|---|
| 9 | Speaks fluently with only rare repetition or self-correction; hesitation is content-related, not word-searching; fully coherent discourse; develops topics fully and appropriately |
| 8 | Fluent with only occasional repetition or self-correction; hesitation usually content-related and only rarely language-searching; develops topics coherently and appropriately |
| 7 | Speaks at length without noticeable effort or loss of coherence; may show language-related hesitation at times, or some repetition or self-correction; uses a range of connectives and discourse markers with some flexibility |
| 6 | Willing to speak at length but may lose coherence due to occasional repetition, self-correction, or hesitation; uses connectives and discourse markers but not always appropriately |
| 5 | Usually maintains flow but uses repetition, self-correction, or slow speech to keep going; may over-use certain connectives; produces simple speech fluently but more complex communication causes fluency problems |
| 4 | Cannot always maintain fluency; produces simple speech fluently but complex communication causes hesitation; limited ability to link simple sentences |
| 3 | Limited ability to sustain speech; frequently speaks in short utterances |
Lexical Resource (LR)
| Band | Descriptor |
|---|---|
| 9 | Uses vocabulary with full flexibility and precision across all topics; uses idiomatic language naturally and accurately |
| 8 | Wide vocabulary readily and flexibly used across topics; uses less common and idiomatic vocabulary with some awareness of style and collocation (with some inappropriate choices); paraphrases effectively |
| 7 | Vocabulary resource flexibly used across topics; uses some less common and idiomatic vocabulary with some awareness of style and collocation; paraphrases effectively |
| 6 | Wide enough vocabulary to discuss topics at length and make meaning clear despite some inappropriacies; generally paraphrases successfully |
| 5 | Manages familiar and unfamiliar topics but uses vocabulary with limited flexibility; attempts paraphrase but with mixed success |
| 4 | Able to discuss familiar topics but can only convey basic meaning on unfamiliar topics; frequent word choice errors; rarely attempts paraphrase |
| 3 | Uses simple vocabulary to convey personal information; insufficient vocabulary for less familiar topics |
Grammatical Range and Accuracy (GRA)
| Band | Descriptor |
|---|---|
| 9 | Full range of structures naturally and appropriately used; consistently accurate structures apart from slips characteristic of native speakers |
| 8 | Wide range of structures flexibly used; majority of utterances error-free; only very occasional inappropriacies or basic/non-systematic errors |
| 7 | Range of complex structures with some flexibility; frequently produces error-free sentences, though some grammatical mistakes persist |
| 6 | Mix of simple and complex structures but with limited flexibility; frequent mistakes with complex structures, though these rarely cause comprehension difficulties |
| 5 | Basic sentence forms with reasonable accuracy; limited range of more complex structures, usually containing errors that may cause some comprehension difficulties |
| 4 | Basic sentence forms and some correct simple sentences; subordinate structures rare; frequent errors may lead to misunderstanding |
| 3 | Attempts basic sentence forms with limited success or relies on memorised utterances; many errors except in memorised expressions |
Pronunciation (P)
| Band | Descriptor |
|---|---|
| 9 | Full range of phonological features with precision and subtlety; effortless to understand |
| 8 | Wide range of phonological features; easy to understand throughout; L1 accent has minimal effect on intelligibility |
| 7 | Shows features of Band 6 and some of Band 8; easy to understand throughout; some features of native-speaker rhythm and intonation |
| 6 | Uses a range of phonological features but control is variable; some effective use of features but not sustained; generally understood but mispronunciation of individual words or sounds reduces clarity at times |
| 5 | Shows features of Band 4 and some of Band 6; accent clearly noticeable; mispronunciations cause some listener difficulty |
| 4 | Limited range of phonological features; attempts to control features with inconsistent success; frequent mispronunciations cause some difficulty for the listener |
| 3 | Shows features of Band 2 and some of Band 4; frequently difficult to understand |
Workflow
Step 1 — Identify what the user needs
Determine:
- Which part are they working on? (1 / 2 / 3 / all)
- Do they want mock questions, a sample answer, or feedback on their own answer?
- Do they have a specific topic or difficulty?
Step 2 — Run the mock or analyse the response
If running a mock:
- For Part 1: ask 3–5 topic questions; after each answer, give brief feedback or save feedback for a summary at the end.
- For Part 2: give a cue card with 4 bullet points; allow the user to write a 1–2 minute response; prompt with "Have you covered all four bullet points?".
- For Part 3: ask 3–4 abstract, analytical follow-up questions linked to the Part 2 topic.
If analysing a transcript:
- Assess naturalness and flow even from written text (hesitations, sentence starters, self-corrections if visible).
- Note vocabulary range and any errors in grammar.
- Estimate pronunciation issues from phonetically unusual constructions where possible, and note what you cannot assess from text.
Step 3 — Score and feedback
Report a band estimate for each criterion and an overall estimated band:
Fluency & Coherence: X / 9
Lexical Resource: X / 9
Grammatical Range: X / 9
Pronunciation: X / 9 (estimated / not assessable from text)
Estimated overall: X.X / 9
For each criterion below 7, give 2–3 specific, actionable improvements with examples drawn from the user's actual response.
Part-specific guidance
Part 1 — Short answers
- Aim for 2–4 sentences per answer: direct answer → reason or example → optional elaboration.
- Avoid one-word answers ("Yes." / "No.") — they give the examiner nothing to assess.
- Avoid over-long answers — Part 1 is meant to be conversational.
Example prompt: "Do you enjoy cooking?"
- Weak: "Yes, I like it."
- Strong: "Yes, I really enjoy cooking, especially on weekends when I have more time. I find it quite relaxing to experiment with new recipes — last month I tried making Vietnamese pho for the first time."
Part 2 — Long turn (cue card)
Structure for the 1–2 minute talk:
- Who / What / Where / When — set the scene quickly (first 20 seconds)
- Description — develop 2–3 points from the bullet list
- How you felt / why it matters — personal reaction or significance
Tips:
- Use the 1-minute prep time to jot 3–4 keywords, not full sentences.
- If you finish early, add a brief reflection ("Looking back, I think this was important because…").
- Do not stop abruptly — the examiner will signal when to finish.
Part 3 — Abstract discussion
- Give structured answers: position → reason → example → implication.
- Use hedging language for opinions: "I would argue that…", "It seems to me that…", "There's a strong case for…"
- Show awareness of different perspectives: "While some people might say X, I think Y because…"
- Part 3 rewards analytical depth — don't just repeat Part 2 ideas.
Key language for speaking
Buying time (natural, not mechanical)
- "That's an interesting question…"
- "Let me think about that for a moment…"
- "It's hard to say exactly, but…"
- "I've never really thought about it that way, but I suppose…"
Avoid overusing "um" and "uh" — they reduce Fluency scores.
Extending answers
- "What I mean by that is…"
- "To give you an example…"
- "This is particularly true when…"
- "In fact, I'd go so far as to say…"
Expressing nuance (important for Band 7+)
- "It really depends on the situation."
- "There are arguments on both sides."
- "While X is generally true, there are exceptions when…"
- "The picture is more complicated than it first appears."
Useful idioms (use naturally, not forced)
- "It goes without saying that…"
- "On balance, I think…"
- "All things considered…"
- "At the end of the day…"
Common mistakes to flag
- One-word or one-sentence answers in Part 1 — always extend with a reason or example.
- Stopping at exactly 1 minute in Part 2 — it is better to fill the full 2 minutes; practise expanding.
- Overusing "because" as the only connective — vary with "due to", "as a result of", "given that", "since".
- Memorised answers — examiners are trained to spot them; they may redirect questions if an answer sounds scripted.
- Saying "I don't know" — always attempt an answer; speculate if unsure ("I'm not entirely sure, but I imagine…").
- Staying in simple tense structures — Band 7+ requires a range of tenses, conditionals, modals, and passive forms.
Output format
When giving mock questions:
- List the questions clearly (Part label + question).
- Wait for the user's response before giving feedback, unless they ask for a sample answer first.
When giving sample answers:
- Provide a realistic Band 7–8 answer.
- Annotate 2–3 features that make it effective (vocabulary choice, structure, grammar).
When scoring a transcript:
- Quote the user's response where relevant.
- Provide the scored table.
- Give a revised version of 1–2 weak sentences to illustrate improvements.
How can the creator link this skill?
Add the canonical catalog link to the repository README so users can inspect current installs and available audits. The publishing guide covers the complete discovery path.
<a href="https://skillzs.dev/skills/jurgendn/agent-skills/ielts-speaking-coach">View ielts-speaking-coach on skillZs</a>