name

agent-interview-coach

description

Set up, run, customize, diagnose, or package a local WeChat AI mock interview coach that reads resume/project materials, generates AI interview background notes, and trains AI/Agent/RAG/MCP/LLM engineering interview answers. Use when the user asks for mobile mock interview practice, WeChat interview bots, resume-based AI interview coaching, dynamic follow-up interviewers, or GitHub packaging of this coach.

Agent Interview Coach

Use this skill to help a user run or maintain the local WeChat interview coach.

Safety

Never print API keys, WeChat tokens, session files, or credential JSON.
Never commit .env, sessions.json, interview_corpus_cache.json, reviews/, answers/, logs, credentials, or real resume files.
Before packaging or publishing, inspect .gitignore and run a secret scan with rg "sk-|token|OPENAI_API_KEY|bot_token|Authorization".

Standard Workflow

Locate the project root.
Check Python dependencies: python -m pip install -r requirements.txt.
Check .env, but do not print secrets.
Prefer the installer on Windows: powershell -ExecutionPolicy Bypass -File .\install_windows.ps1.
Verify model from app/: python smoke_test.py --model.
Verify WeChat from app/: python smoke_test.py --wechat.
If WeChat is unavailable, run CLI fallback from app/: python cli_chat.py.
Start on Windows from app/: powershell -ExecutionPolicy Bypass -File .\start_bot.ps1.
Check logs from app/: powershell -ExecutionPolicy Bypass -File .\status_bot.ps1.
Run full diagnostics from repo root: powershell -ExecutionPolicy Bypass -File .\doctor.ps1.

Resume Material Workflow

Tell users to either:

Put .docx/.pdf/.md/.txt files in the bot's resume_materials/ folder and send /导入资料 <folder>.
Or send /导入资料 <absolute-folder-path> for an existing resume/project material folder.

Then send /生成背景 to create AI面试背景材料.generated.md.

WeChat Commands

/资料入口: show the local folder for resume materials.
/导入资料路径: switch material folder and rebuild corpus.
/生成背景: generate candidate-specific interview background.
/模式教练|技术面|拷打|只面试: switch training style.
开始电话面|开始一面|开始二面|开始HR面: switch interview stage.
/今日弱点: show tracked weaknesses.
/复盘: write a Markdown review.
/标准答案: save the latest improved answer.
/刷新: reread materials.
/重置: clear session state.

Troubleshooting

If model auth fails, check .env values and whether the provider supports MODEL_API_STYLE=chat.
If WeChat polling fails, rerun wechat-clawbot-cc setup.
If /刷新 reads the wrong files, check RESUME_SOURCE_DIR.
If output is garbled in PowerShell, set $env:PYTHONIOENCODING='utf-8' for terminal tests; the Python files remain UTF-8.
If the user wants to test without WeChat, use python cli_chat.py.

Agent Interview Coach

Use this skill to help a user run or maintain the local WeChat interview coach.

Safety

Never print API keys, WeChat tokens, session files, or credential JSON.

Never commit .env, sessions.json, interview_corpus_cache.json, reviews/, answers/, logs, credentials, or real resume files.

Before packaging or publishing, inspect .gitignore and run a secret scan with rg "sk-|token|OPENAI_API_KEY|bot_token|Authorization".

Standard Workflow

Locate the project root.

Check Python dependencies: python -m pip install -r requirements.txt.

Check .env, but do not print secrets.

Prefer the installer on Windows: powershell -ExecutionPolicy Bypass -File .\install_windows.ps1.

Verify model from app/: python smoke_test.py --model.

Verify WeChat from app/: python smoke_test.py --wechat.

If WeChat is unavailable, run CLI fallback from app/: python cli_chat.py.

Start on Windows from app/: powershell -ExecutionPolicy Bypass -File .\start_bot.ps1.

Check logs from app/: powershell -ExecutionPolicy Bypass -File .\status_bot.ps1.

Run full diagnostics from repo root: powershell -ExecutionPolicy Bypass -File .\doctor.ps1.

Resume Material Workflow

Tell users to either:

Put .docx/.pdf/.md/.txt files in the bot's resume_materials/ folder and send /导入资料 <folder>.

Or send /导入资料 <absolute-folder-path> for an existing resume/project material folder.

Then send /生成背景 to create AI面试背景材料.generated.md.

WeChat Commands

/资料入口: show the local folder for resume materials.

/导入资料路径: switch material folder and rebuild corpus.

/生成背景: generate candidate-specific interview background.

/模式教练|技术面|拷打|只面试: switch training style.

开始电话面|开始一面|开始二面|开始HR面: switch interview stage.

/今日弱点: show tracked weaknesses.

/复盘: write a Markdown review.

/标准答案: save the latest improved answer.

/刷新: reread materials.

/重置: clear session state.

Troubleshooting

If model auth fails, check .env values and whether the provider supports MODEL_API_STYLE=chat.

If WeChat polling fails, rerun wechat-clawbot-cc setup.

If /刷新 reads the wrong files, check RESUME_SOURCE_DIR.

If output is garbled in PowerShell, set $env:PYTHONIOENCODING='utf-8' for terminal tests; the Python files remain UTF-8.

If the user wants to test without WeChat, use python cli_chat.py.