Help us improve
Share bugs, ideas, or general feedback.
From sharepoint
Design, repair, and roll out formulas in SharePoint-hosted workbooks with connector-aware retrieval, validation, and upload discipline. Use when the user wants to add a formula column, fix a broken formula, choose between a fill-down formula and a spill formula, build a lookup or filter formula, or reuse workbook logic safely.
npx claudepluginhub robinebers/converted-plugins --plugin sharepointHow this skill is triggered — by the user, by Claude, or both
Slash command
/sharepoint:sharepoint-spreadsheet-formula-builderThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this skill when the formula itself is the task and the workbook lives in SharePoint.
Guides Next.js Cache Components and Partial Prerendering (PPR): 'use cache' directives, cacheLife(), cacheTag(), revalidateTag() for caching, invalidation, static/dynamic optimization. Auto-activates on cacheComponents: true.
Creates, reads, edits, and analyzes .docx files using docx-js for new documents, pandoc for text extraction, Python scripts for XML unpacking/validation/changes, and LibreOffice for conversions.
Share bugs, ideas, or general feedback.
Use this skill when the formula itself is the task and the workbook lives in SharePoint.
This workflow depends on the Microsoft SharePoint connector for file discovery and binary transfer, not formula-aware editing. The relevant connector actions on the exposed Codex surface are get_site, list_site_drives, search, list_folder_items, fetch, update_file, and upload_file.
In Codex, use those direct Microsoft SharePoint app tools rather than generic MCP resource-listing flows. The backend wrapper routes connector actions through tool calls, not through user-facing SharePoint resource enumeration.
Read ./references/formula-patterns.md before drafting the first formula. The point is to refresh exact Excel syntax, formula-shape choices, and SharePoint-specific rollout risks before editing the workbook.
get_site when you need to confirm the canonical site before path-based operationslist_site_drives when the site is known but the right library is notsearch(query="...") when you have keywordssearch(query=None, hostname=..., site_path=..., folder_path=...) to browse a known site or folderlist_folder_items when the exact folder path is already knownurl returned by keyword search or browse results as the input to fetchfetch once for extracted content to identify sheets and likely target rangesfetch(download_raw_file=true) for the actual .xlsx bytes before editing formulasLET when repeated subexpressions make the formula hard to read or maintainLAMBDA or a named formula only when the logic is conceptually reusable and the workbook environment supports itupdate_file when replacing an existing workbook at a known pathupload_file only when creating a new workbook or when the workflow genuinely requires upload semantics.xlsx, then goes back through the SharePoint file-update path.fetch, but they are fallback inputs rather than the preferred primary path.search only returns text-friendly document types and enforces the connector's file-size gate. If keyword search is not the right tool, switch to site-scoped browse mode instead of using a vague fallback.XLOOKUP, FILTER, LET, and LAMBDA as workbook-compatibility choices, not connector features. Use them when the target workbook is expected to open in Microsoft 365-era Excel or Excel for the web, and call out compatibility risk otherwise.IF, IFERROR, SUMIFS, COUNTIFS, INDEX/MATCH, and exact-match VLOOKUP(FALSE) over newer dynamic-array constructs.itemNotFound, re-check the root-relative path from SharePoint metadata instead of guessing from the browser URL../references/formula-patterns.md.