Hk Intership

HK Intership Workspace Router

This skill is the entry point for the G:\Wyatt\HK_intership workspace.

Canonical local skills live at:

• G:\Wyatt\HK_intership\.agent\skills\research-workflow
• G:\Wyatt\HK_intership\.agent\skills\research-analysis-subagent
• G:\Wyatt\HK_intership\.agent\skills\expert-survey-report

Repo helper tools live at:

• G:\Wyatt\HK_intership\.agent\tools\verify-research-runtime.ps1
• G:\Wyatt\HK_intership\.agent\tools\research-runtime.ps1
• G:\Wyatt\HK_intership\.agent\tools\install-research-runtime.ps1

Use this skill to route work inside the HK_intership workspace:

1. If the user wants an end-to-end paper workflow, use research-workflow. Typical requests include:
   • generating a research report from paper links
   • batch-downloading papers
   • converting PDFs to Markdown
   • merging/exporting a report to Word
   • typesetting/exporting the final merged report to PDF via lovstudio-md2pdf
2. If the user wants an expert-survey-style final report that is theme-organized rather than paper-by-paper, use expert-survey-report. Typical requests include:
   • expert survey
   • consulting-grade review
   • 按主题归类，不要逐篇总结
   • 不要文献卡片感
   • 先立研究对象、主线和中心判断
3. If the user wants to process one paper directory and write Analysis_Detail.md, use research-analysis-subagent.
4. Before running the research pipeline, verify the runtime with: powershell -ExecutionPolicy Bypass -File "G:\Wyatt\HK_intership\.agent\tools\verify-research-runtime.ps1"
5. Keep G:\Wyatt\HK_intership\.agent\skills\... as the source of truth for scripts, prompts, and assets.
6. Prefer G:\ paths for outputs and runtime artifacts. Avoid saving to C:\ unless the user explicitly requests it.

Workspace landmarks:

• downloads\ holds task outputs and paper batches.
• docs\superpowers\specs\ and docs\superpowers\plans\ hold prior design and implementation notes.
• .agent\skills\ contains canonical skill instructions.
• .agent\tools\ contains runtime bootstrap and verification scripts.