← Back to Skills Marketplace
zhiyicom

English Oral Tutor

by zhiyicom · GitHub ↗ · v1.0.4 · MIT-0
cross-platform ⚠ suspicious
136
Downloads
0
Stars
0
Active Installs
5
Versions
Install in OpenClaw
/install english-oral-tutor
Description
Provides voice-based English speaking lessons and conversation practice for Chinese Grade 7 students, including pronunciation correction and mic setup help.
README (SKILL.md)

English Oral Tutor

Voice-based English conversation teacher for a Chinese middle school student.

Student Profile

  • Age: 13-14 (Grade 7, first year of middle school)
  • Location: Beijing, China
  • Level: Pre-intermediate (CEFR B1 / Chinese middle school English)
  • Vocabulary constraint: Keep vocabulary within Cambridge B1 level and Chinese middle school English syllabus
  • Needs: Spoken English practice, conversation fluency, grammar accuracy
  • Name: (Ask at first meeting)

Teaching Rules

  1. Speak English only — Never translate to Chinese during lesson
  2. Use B1-level vocabulary — No words beyond Cambridge B1 or Chinese middle school English scope
  3. Be encouraging — Praise effort, not just correctness
  4. Correct grammar immediately — Gently point out errors: "I think you meant..." or "Actually, we say it like this..."
  5. Keep it fun — Age-appropriate topics, games, things the student cares about
  6. No emoji in speech — Avoid reading emoji aloud
  7. Active topics only — Never repeat topics already discussed (check conversation-history.md first)
  8. Go deeper — Ask follow-up questions to push conversation forward, don't stay on surface level

Voice Settings (TTS) — Current Config

  • Provider: Microsoft Edge TTS
  • Voice: en-US-JennyNeural
  • Rate: +25%
  • Pitch: 0%
  • Config: messages.tts.provider: microsoft

Conversation Management

Topic Tracking

  • All discussed topics go into references/conversation-history.md
  • Before starting a new topic, check if it was already discussed
  • If discussed before: continue from where left off, introduce new sub-angle
  • Never repeat specific discussion points already covered

Topic Library

  • See references/topic-library.md for collected Cambridge B1 / CEFR口语考试 topics
  • Pick topics appropriate for 13-14 year old
  • Mix familiar topics with slightly challenging new ones

Grammar Correction Rules

  • When correcting: be gentle, never make the student feel embarrassed
  • Say: "Great try! Just one small thing — we usually say..." rather than "That's wrong"
  • After correction, ask the student to say the correct version aloud
  • Log the error type in references/conversation-history.md under the topic

Lesson Structure

  1. Warm-up (2-3 min) — Greet and ask one easy question about their day/week
  2. Main activity (10-15 min) — Introduce topic, teach 2-3 key vocabulary words, have conversation
  3. Grammar check — Note errors and correct them during conversation
  4. Wrap-up (2-3 min) — Summarize what was practiced, assign mini-practice, say goodbye

Voice Output Rules

  • Do NOT read emoji aloud
  • Keep replies under 1500 characters for TTS
  • Use natural pauses, conversational tone
  • For emojis in text: describe verbally (e.g., "smiling face" instead of reading "😊")
  • Ask follow-up questions to keep conversation going

Technical Fixes

Permissions-Policy Microphone (Browser Speech Recognition)

If the microphone button in Control UI doesn't work, the browser's Permissions-Policy header may be blocking it.

Fix: Run scripts/fix-microphone.ps1 or manually patch the gateway CLI:

copy "C:\Users\samuel\.openclaw\workspace-english-teacher\gateway-cli-Dsd9gHBa.js" "C:\Users\samuel\AppData\Roaming\
pm\
ode_modules\openclaw\dist\gateway-cli-Dsd9gHBa.js"
openclaw gateway restart

See references/technical-fix.md for full procedure.

Alt Key Microphone Shortcut

Adds Alt key to toggle microphone when the chat input is empty and focused. Without this fix, the microphone only works via mouse click on the button.

Fix: Run scripts/fix-microphone-shortcut.ps1 — then refresh the Control UI page.

  • Already idempotent: safe to run even if already patched
  • After OpenClaw updates, re-run this script to restore the shortcut

TTS Voice Speed

Adjust rate: openclaw config set messages.tts.edge.rate "+25%" then restart gateway.

Usage Guidance
What to consider before installing or following this skill: - The tutoring content is coherent, but the technical fixes instruct you to overwrite and patch OpenClaw's installed files (node_modules) and to run PowerShell scripts. That can change system-wide behavior and requires write/admin access — do not run these steps blindly. - The SKILL.md references scripts (scripts/fix-microphone*.ps1) that are not included in the package. Ask the author for the actual scripts and inspect their contents before running them. - Hardcoded file paths (C:\Users\samuel\...) are suspicious: they may not match your environment and suggest the instructions were copied from a single developer's machine. Verify and adapt paths carefully. - If you need the mic fixes, prefer safe alternatives: ask for patch diffs rather than binary copies, back up original files first, or test in an isolated environment (VM) before modifying your primary system. - Consider requesting provenance: who published this skill, where did the scripts/patches come from, and can they provide signed releases or source code for review? - If you must apply fixes, make incremental backups, review any scripts line-by-line, avoid running anything that contacts unknown external endpoints, and keep the skill disabled from autonomous runs until you trust it. Given the mixture of reasonable tutoring content and risky system-modifying instructions, treat this skill with caution and seek the missing scripts/source before proceeding.
Capability Analysis
Type: OpenClaw Skill Name: english-oral-tutor Version: 1.0.4 The skill contains instructions (SKILL.md) and technical documentation (references/technical-fix.md) that direct the AI agent to modify the OpenClaw application's core source code within the local 'node_modules' directory. Specifically, it provides code to patch 'gateway-cli-Dsd9gHBa.js' to alter security headers (Permissions-Policy) and injects JavaScript into 'index-sc9GrLxu.js' to add keyboard shortcuts. While these are framed as functional fixes for microphone issues, the practice of self-patching the host application via PowerShell scripts and manual file overrides is a high-risk behavior that could be leveraged for persistence or to introduce vulnerabilities.
Capability Assessment
Purpose & Capability
The skill's voice tutoring purpose justifies helping with microphone and TTS setup, so the technical-fix content is related. However, the SKILL.md directs edits to global OpenClaw files (AppData\Roaming\npm\node_modules\openclaw\dist\...) and to a workspace backup path (C:\Users\samuel\...), which is broader and more powerful than a typical tutor skill would need. Hardcoded username paths ('samuel') and instructions to overwrite files in a global package are disproportionate and unusual for a simple lesson skill.
Instruction Scope
Instructions explicitly tell the user/agent to copy and patch gateway-cli JavaScript inside an installed openclaw package, run PowerShell scripts, run npm pack and edit extracted files, and insert JS snippets into a bundled UI asset. Those runtime instructions involve reading and writing system files, require elevated privileges in some environments, and can affect other OpenClaw functionality. The skill references scripts (scripts/fix-microphone.ps1, scripts/fix-microphone-shortcut.ps1) that are not present in the file manifest, increasing risk: the agent/user would need to obtain or run external scripts not included with the skill.
Install Mechanism
This is an instruction-only skill with no install spec or remote downloads, which reduces supply-chain risk. The risk arises from the instructions (file modifications) rather than an installer or remote archive.
Credentials
The skill requests no environment variables or credentials (which is good). However, it asks to modify files under system/global installation paths (node_modules) and the user's home directory — operations that require filesystem write access and sometimes elevated privileges. No secrets are requested, but filesystem-level changes are high-privilege relative to a tutoring task and should be justified and inspected.
Persistence & Privilege
The skill is not marked always:true, and autonomous invocation is allowed (the platform default). The central concern is that the instructions modify another package's installed files (OpenClaw's dist files), which changes system-wide behavior beyond this skill's scope. A skill directing edits to other components is a privilege escalation risk and should be treated cautiously.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install english-oral-tutor
  3. After installation, invoke the skill by name or use /english-oral-tutor
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.4
更新摘要为本skill描述的中文版本
v1.0.3
添加中文说明:本skill专为OpenClaw内置Control UI设计,支持语音交互(麦克风快捷键ArrowDown)、TTS自动播放、语法纠错、话题追踪等功能,适合CEFR B1级别的英语口语练习。
v1.0.2
Added: Supports voice interaction via microphone (ArrowDown shortcut) and TTS auto-play for full voice-based conversation practice.
v1.0.1
Updated description: This skill is designed specifically for OpenClaw's built-in Control UI (web interface). It includes TTS auto-play via injected JavaScript, microphone shortcut key (ArrowDown), and all necessary technical patches for the Control UI environment.
v1.0.0
Initial release: Voice-based English tutoring for Chinese Grade 7 student (CEFR B1). Features: TTS auto-play, microphone shortcut (ArrowDown), grammar correction rules, topic library, conversation history tracking.
Metadata
Slug english-oral-tutor
Version 1.0.4
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 5
Frequently Asked Questions

What is English Oral Tutor?

Provides voice-based English speaking lessons and conversation practice for Chinese Grade 7 students, including pronunciation correction and mic setup help. It is an AI Agent Skill for Claude Code / OpenClaw, with 136 downloads so far.

How do I install English Oral Tutor?

Run "/install english-oral-tutor" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is English Oral Tutor free?

Yes, English Oral Tutor is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does English Oral Tutor support?

English Oral Tutor is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created English Oral Tutor?

It is built and maintained by zhiyicom (@zhiyicom); the current version is v1.0.4.

💬 Comments