Jobo16/ielts-buddy / skills/ielts-speaking-coach
ielts-speaking-coach
Verified进行雅思 Part 1、Part 2、Part 3 逐题陪练、表达反馈和真实经历素材整理,或建立一组已知题目的故事覆盖图。文字与音频证据分别处理。
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 1.7 KB |
| agents/openai.yaml | config | 260 B |
| manifest.json | config | 214 B |
| references/api-access.md | markdown | 1.1 KB |
| references/browser-routes.md | markdown | 587 B |
| references/skill-taxonomy.json | config | 1.5 KB |
| references/speaking-coach/WORKFLOW.md | markdown | 5.1 KB |
| references/speaking-coverage-map/WORKFLOW.md | markdown | 3.3 KB |
| references/speaking-topic-builder/WORKFLOW.md | markdown | 3.3 KB |
| references/speaking-vocabulary-handoff.md | markdown | 1.3 KB |
| scripts/simple_docx.py | script | 9.0 KB |
| scripts/speaking-coach/create_speaking_report_docx.py | script | 3.7 KB |
| scripts/speaking-coach/validate_speaking_report_docx.py | script | 2.3 KB |
references/speaking-topic-builder/WORKFLOW.md
View on GitHub# Speaking Topic Builder
> 当当前学习任务需要本方法时按需读取。脚本命令中的 `scripts/` 相对于本 Skill 根目录;执行时解析为安装目录中的绝对路径。
默认面向中国雅思学习者,用简体中文解释选择理由、练习方法和复盘要求。题目、答案骨架、英文表达块保留英文。
## Core Rule
本地 Agent 负责选题、串题、素材组织和练习策略。IELTS Buddy 只提供题库数据、已有材料、进度和可打开链接。不要让服务端自动匹配 suitable questions。
## Inputs
Use the smallest useful set:
- learner goal, target band, exam date, weak speaking part;
- learner's core story, experience, topic preference, or existing answer;
- IELTS Buddy practice taxonomy/search/read data when available;
- existing speaking materials from IELTS Buddy when the user wants to update old material.
If exact questions are unavailable, use the learner-provided cue cards or return the browser practice entry. Save only a title and real core story when requested.
## Local Workflow
1. Identify the practice target:
- Part 1: natural short answers and follow-up range.
- Part 2: one reusable story with cue-card coverage.
- Part 3: opinion development, examples, comparison, and abstraction.
2. Read available data:
- Use `ielts_practice_list_taxonomy`, `ielts_practice_search_parts`, or `ielts_practice_read_part` when the learner needs question-bank choices.
- Use `ielts_speaking_materials_search` and `ielts_speaking_materials_read` only to inspect owned saved materials.
3. Select exact questions locally.
- Prefer questions that reuse the learner's real experiences.
- Avoid grouping unrelated topics just because vocabulary overlaps.
- Keep the set small: 3-5 Part 1 questions, 1 Part 2 cue card plus 3 follow-ups, or 4-6 Part 3 questions.
4. Build the material:
- `core_story`: one realistic personal story or opinion base.
- `本地题目映射`: exact IELTS Buddy question/session/resource links or IDs.
- `answer_angles`: concise Chinese notes for what to say.
- `reusable_chunks`: 5-8 English chunks.
- `avoid`: 2-4 common weak expressions or logic traps.
- `practice_order`: easy-to-hard sequence.
5. Coach the first answer locally or hand off to `../speaking-coach/WORKFLOW.md`.
6. Persist only if the Agent API tool accepts caller-provided fields without AI matching. If the manifest says omitted links trigger AI matching, supply exact links; if exact links are impossible, skip persistence and return the local material plus browser route.
## Response Shape
```text
本次口语素材主题:<theme>
适合题型:<Part 1/2/3>
为什么这样选:<one evidence-based reason>
题目顺序:
1. <question + link/id>
2. ...
可复用故事线:<Chinese outline>
可复用表达:<English chunks>
先练这一题:<question>
回答后我会按 speaking-coach 做本地反馈。
```
## Agent API Write Policy
Use `ielts_speaking_materials_create` or `ielts_speaking_materials_update` only when all teaching decisions are already made locally and the payload contains explicit question links/IDs plus caller-provided material fields.
Do not call these tools with missing question links, vague matching instructions, or a request for the server to choose questions.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |