OpenSenseNova/SenseNova-Skills — reverse-engineered prompt

Build me a Python based SenseNova skills pack for an AI office assistant. I want it to work with Agent Skills compatible agents like OpenClaw or Hermes, where each skill can be dropped into the agent skills folder and discovered from its SKILL.md file.

The assistant should help with real office work, not just demos. It should be able to generate and understand images, make polished infographics from prompts or reports, imitate a reference image with new content, create resume images, generate PowerPoint decks from documents or a short brief, analyze Excel files, and support deeper research style workflows.

Please include clear setup docs, example prompts, environment checks for missing dependencies or SenseNova API settings, and a simple way to install the skills into the right local agent folder. Keep the skills modular so they can be used alone or chained together into full workflows. Look up the current Agent Skills and SenseNova docs online if you need to.