SenseNova Skills

SenseNova Skills is an open-source collection of modular skills for AI Agents, built around the agent skills convention. Rather than operating as a Standalone chatbot, it packages office workflows so an Agent can move from understanding source material to producing a finished deliverable such as a slIDE deck, spreadsheet analysis, research report, or infographic. The project comes from SenseTime, and its core idea is to close the gap between a General model’s reasoning ability and the prACTical steps required to complete real knowledge work.

What it does

SenseNova Skills provides reusable skill components that an agent can discover and invoke. Each skill declares when it should be used, what it can do, and how it should execute. The library groups capabilities around multimodal understanding, interactive editing, and multimodal generation. In practice, this means an agent can read documents, tables, screenshots, and images; combine Information from multiple sources; run analysis; refine a draft through conversation; and export usable files. The project also includes research and search skills for academic, developer, and social sources, with some integrations requiring additional API keys or platform access.

Key capabilities

• presentation Workflows: Generate slide decks through standard, creative, or quick modes, with options for editable HTML-based ouTPUt or image-based creative pages.
• Spreadsheet analysis: Read multiple sheets, clean and filter data, aggregate across tables, handle large files, and export results to Excel or CSV.
• Deep research: Plan research before searching, gather evidence from sources, reconcile conflicts, and produce structured reports with citations.
• Infographic generation: Turn complex material into visual summaries using multiple layouts and styles, with quality checks and rAnking.
• multimodal input handling: Interpret tables, charts, screenshots, documents, and images, then convert them into structured data or downstream task inputs.
• Agent compatibility: designed for runtimes and tools that support Agent Skills or MCP-style integrations, including developer-oriented environments.

Who it is for

SenseNova Skills is AImed at developers and teams building AI office agents, as well as analysts, consultants, researchers, educators, and Operations staff who need repeatable outputs rather than one-off suggestions. It is especially relevant for users who want open, inspectable workflows and the ability to combine skills into longer chains, such as data analysis followed by industry research and a presentation. Nontechnical users may prefer a hosted product that bundles the skills, while teams with engineering resources can integrate the open-source components into their own agents.

How to get started

1. Review the open-source repository and confirm that your agent runtime supports the Agent Skills fORMat.
2. Obtain the required SenseNova API credentials for the underlying models.
3. Install or clone the skills into a compatible environment, then configure API keys and dependencies.
4. Start with a single skill, such as spreadsheet analysis or presentation generation, before combining multiple skills into a longer workflow.

Pricing and access caveat

The skill code is released under an open-source license, but model calls may require a SenseNova API key and incur token costs. A hosted SenseTime office product may offer trial access, and availability, quotas, and pricing can vary by region and account. Some search skills may need separate third-party API keys or authentication. Users should verify current access terms before relying on the toolkit for production work.