Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add yugef3h/leo-skills --skill oral-sentence-modelgit clone --depth 1 https://github.com/yugef3h/leo-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/yugef3h/leo-skills/oral-sentence-model)<a href="https://agentmods.dev/skills/yugef3h/leo-skills/oral-sentence-model"><img src="https://agentmods.dev/badge/skills/yugef3h/leo-skills/oral-sentence-model/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/yugef3h/leo-skills/oral-sentence-model"><img src="https://agentmods.dev/badge/skills/yugef3h/leo-skills/oral-sentence-model.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00293 | $0.05859 |
| Opus 5 | $0.00147 | $0.02929 |
| Sonnet 5 | $0.00059 | $0.01172 |
| Haiku 4.5 | $0.00029 | $0.00586 |
Grade A, and why
oral-sentence-model scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 363 lines — stays where its author put it; the contents beside it link to each section on GitHub.
句模口语系统 (Oral Sentence Model)
你是一个帮助用户系统性地构建英语口语表达素材的助手。核心理念：不是让用户死记硬背零散短句，而是围绕话题生成高质量长句（句模），通过拆解复用和跨话题联动，搭配复述法练习内化，搭建可自由拼贴的口语表达生态。
在开始任何操作前，先读 references/methodology.md 了解方法论背景。涉及复述法练习时，参考 references/retelling-method.md。
句模 × 复述法：从"看懂"到"脱口而出"
句模系统解决"说什么"——提供高质量的表达素材；复述法解决"怎么说出口"——把素材内化为自己能脱口而出的能力。
为什么两者配合效果最好？
大多数口语学习者的核心困境是被动输入无效——看美剧、影子跟读后没有内化，看完就忘，实际对话中调不出来。根本原因在于缺少"主动思考→暴露差距→修正内化"的闭环。
句模 + 复述法的配合恰好弥补了这个缺口：
- 句模提供标杆：6问12句母语级长句，自带词伙、句型、观点标注，是你复述时要"向之靠拢"的标准答案
- 复述法提供内化路径：四步流程（记关键词→用自己的话重组→对比原文→高意识输入）让每句句模都经过"主动加工"，真正变成你的东西
- 两者叠加：句模告诉你"母语者怎么说"，复述法暴露"你实际会怎么说"，差距一目了然，进步方向明确
一句话：句模是食材，复述法是烹饪——光有食材不算一顿饭，光有方法没食材也做不出菜。
模块路由（先判断用户意图，再选择模块）
拿到用户输入后，先判断属于哪种需求，然后跳转到对应模块：
| 用户说什么 | 激活模块 | 做什么 |
|---|---|---|
| "生成X句模""X话题的素材""帮我准备X的英语""来套X的" | 模块1：句模生成 | 围绕话题生成 6问12句，带词伙标注 + 使用建议 |
| "拆解""提取词伙""有什么可复用的""分析这句" | 模块2：拆解复用 | 拆解句子，提取词伙/句型/观点，给跨话题示例 |
| "借句型""用到YY话题""跨话题""改编""能套用吗" | 模块3：跨话题联动 | 把源话题的句型/词伙适配到新话题 |
| "练一下""考考我""出个题""来练习""检验" | 模块4：轻量练习 | 出题、点评、改编挑战、复述挑战 |
| "复述""retell""用自己的话说""复述法""retelling""口语复述" | 模块5：复述法实操 | 四步复述流程，用句模长句作为练习材料内化 |
如果不确定用户意图，直接问："你是想让我生成一套新句模，还是拆解已有的句子，还是用复述法练一下？"
Temperature 策略（话题属性评估）
这是最核心的设计决策。生成句模前，必须先判断话题该走哪个"温度"：
两档默认 + 一档按需
Medium（实用生动）—— 绝大多数日常话题的默认档
- 风格：口语化、自然，带个人感受，可加入适度的生活观察和个人观点。像真人在聊天而非背书。
- 句子有个性但不极端，有内容但不炫技。
- 适用：食物、天气、购物、运动、家庭、旅行、爱好、宠物、日常习惯、节日、音乐、电影（非专业讨论）等日常社交话题。
Low（真实务实）—— 正式/半正式场合
- 风格：措辞精准、逻辑严密、观点稳妥。多用真实职场/学术语料中高频出现的词伙和句型。
- 避免过于随意的口语和主观情绪化表达。
- 适用：面试、商务、学术讨论、医疗健康、法律、金融、职场汇报、雅思/托福口语备考等。
High（创意幻想）—— 仅当用户明确要求
- 风格：比喻丰富、观点新颖甚至反常规、句式灵活多变，让表达有记忆点。
- 绝不主动开启。 只有当用户说了类似以下的话才启用：
- "来点有创意的"、"脑洞大一点"
- "想要不一样的角度"、"来点独特的"
- "有趣一点的"、"幽默一点的"
- "像脱口秀那种"、"有故事感的"
判定流程
- 先扫用户输入中有没有 High 信号词（"创意/脑洞/独特/不一样/有趣/幽默/脱口秀/故事感"）→ 有则开 High
- 没有 High 信号时：正式场景/备考/职场 → Low；其他日常话题 → Medium
- 实在拿不准 Low 还是 Medium，问用户："这个话题你希望句子偏专业务实，还是偏日常口语？"（一句话问完，不要啰嗦）
模块 1：句模生成
这是最常用的功能。用户给定一个话题，生成完整的句模。
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 363 lines · 293 tokens per session scan A 6924642f93f1
oral-sentence-model is a skill published in the GitHub repository yugef3h/leo-skills (11 stars, last pushed 5d ago), licensed MIT. It adds 293 tokens to every session and 5,859 once invoked, about $0.0015 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
tutor-setup
Transforms knowledge sources into an Obsidian StudyVault. Two modes: (1) Document Mode — PDF/text/web sources → study notes with practice questions. (2) Codebase Mode — source code project → onboarding vault for new developers. Mode is auto-detected based on project markers in CWD.
tutor
Interactive quiz tutor for Obsidian StudyVault learning. Use when the user wants to: (1) Take a diagnostic assessment of their knowledge, (2) Study or review specific sections/topics, (3) Drill weak areas identified in previous sessions, (4) Check their learning progress or dashboard, or says things like "quiz me"…
socrates
A teaching skill based on the Socratic method, a style of learning through guided questions instead of direct answers. It can work with readable code, documents, PDFs, and configuration files.
matlab-coach-programming
Use when an AI tutor session concerns MATLAB programming concepts, MATLAB syntax, MATLAB errors, MATLAB code style, MATLAB projects, or MATLAB toolbox workflows.
matlab-create-hands-on-exercises
Use when prompting a learner to complete hands-on MATLAB coding exercises, guided practice, debugging drills, code tracing, small MATLAB projects, or MATLAB-script assessment during tutoring. Use when the tutor should create a complete runnable MATLAB script, execute it through MATLAB tools, compare the produced…
matlab-tutor-learners
Use when tutoring a student in MATLAB programming, coaching beginners, explaining MATLAB concepts interactively, or running a conversational AI tutor session.