About this skill

What problem it solves

When an agent needs to reliably write meeting notes, structured data, slide decks, or local files into Tencent Docs, scattered tool interfaces, inconsistent document types, and blocking async tasks become common failure points. This skill turns those operations into a routed MCP workflow: pick the right document type by scenario, then create, read, upload, clip, and debug through consistent paths.

Core capabilities and key steps

Default creation: prefer smartcanvas for new documents; write Markdown directly into the mdx parameter of create_smartcanvas_by_mdx.
Structured data: use sheet.* for Excel-like data and smartsheet.* for advanced sheets, resolving sheet_id before further operations.
Content and files: read content with get_content; use manage.* for renaming, moving, deleting, copying, import/export, and permissions.
Web and local content: for URLs, prefer scrape_url plus scrape_progress; for local files, use import_file.sh, manage.async_import, and manage.import_progress.
OCR: call ocr.* for public image URLs; use the provided script path for local images instead of manually handling base64.
Debugging: check error codes first, then verify the live tool schema with mcporter list tencent-docs or mcporter list slide-mcp.

Scope and cautions

This skill is suited for Tencent Docs read/write, upload, web clipping, and presentation generation, not content compliance review. Auth state, VIP access, credits, and async task status can affect outcomes. Treat delete_space_node recursive deletion carefully, and note that create_smartcanvas_by_mdx does not support parent_id.