SharePoint Automation
by ComposioHQSharePoint Automation skill for using Rube MCP to manage SharePoint and OneDrive sites, lists, documents, folders, pages, and search, with setup notes for organizational Microsoft 365 accounts.
This skill scores 72/100, which means it is acceptable for directory listing but best suited for users already comfortable with Rube/Composio and Microsoft 365 integrations. The SKILL.md gives agents enough trigger terms, tool mappings, and parameter examples to improve over a generic prompt, but install decision confidence is limited by the lack of support files, install command, and fuller operational examples in the repository evidence.
- Clear scope for automating SharePoint and OneDrive operations across sites, lists, documents, folders, pages, and search.
- Includes concrete tool names such as `ONE_DRIVE_GET_SITE_DETAILS` and `ONE_DRIVE_LIST_SITE_SUBSITES`, plus key parameter guidance like the required composite `site_id` format.
- Provides setup prerequisites and an important account limitation warning before tool execution.
- Requires the Rube MCP server and active `share_point` plus `one_drive` connections; users without organizational Microsoft 365/Azure AD accounts may not be able to use many tools.
- No install command or support files are provided, so setup and troubleshooting depend mostly on the single SKILL.md and external Composio toolkit docs.
Overview of SharePoint Automation skill
What SharePoint Automation does
SharePoint Automation is a Claude skill for operating SharePoint and OneDrive through the Rube MCP server. It helps an agent browse SharePoint sites, work with lists, manage documents and folders, handle pages, and search content using the share_point and one_drive toolkits. It is most useful when you want an AI assistant to perform structured Microsoft 365 content operations instead of only describing how to do them.
Best fit for SharePoint and document workflows
This SharePoint Automation skill fits teams that already use organizational Microsoft 365, especially operations, legal, HR, finance, project management, and knowledge-management groups that need repeatable document handling. A strong use case is SharePoint Automation for Document Filing: locating the right site or library, creating folders, uploading or moving files, updating list records, and verifying where content ended up.
Key adoption requirement
The main blocker is connectivity, not prompt quality. The skill requires the Rube MCP server at https://rube.app/mcp and active connections for both share_point and one_drive. Many tools depend on organizational Azure AD / Entra ID accounts; personal Microsoft accounts are not supported. Before installing, confirm your environment allows MCP access, Microsoft 365 authentication, and the permissions needed for the target SharePoint sites.
How to Use SharePoint Automation skill
SharePoint Automation install and connection setup
Install the skill from the repository path used by your skills manager. For Claude skill workflows, a typical command is:
npx skills add ComposioHQ/awesome-claude-skills --skill "SharePoint Automation"
Then inspect composio-skills/share-point-automation/SKILL.md; this is the primary source file and there are no separate bundled scripts, rules, or reference folders in the current skill package. In your MCP configuration, connect Rube and authorize the share_point and one_drive toolkits. If the agent reports no active connection, ask it to start connection setup with RUBE_MANAGE_CONNECTIONS before attempting SharePoint actions.
Inputs the skill needs to work reliably
SharePoint tools are sensitive to identifiers and permissions. For site operations, the skill expects a site_id in the Microsoft Graph composite format:
hostname,site-collection-guid,web-guid
For example:
contoso.sharepoint.com,2C712604-1370-44E7-A1F5-426573FDA80A,2D2244C3-251A-49EA-93A8-39E1C3A060FE
For document and list workflows, provide the site, library or list name, target folder path, file names, metadata fields, and whether the agent should create, update, move, search, or only report. If you do not know the site ID, start with a discovery task such as listing available sites, getting site details, or enumerating subsites.
Strong SharePoint Automation usage prompts
A weak prompt is: “File these documents in SharePoint.” It leaves the agent guessing about site, library, naming rules, and permissions.
A stronger prompt is:
“Use SharePoint Automation to file the uploaded vendor contracts. First confirm the active share_point and one_drive connections. Locate the Finance site, then the Contracts document library. Create a folder path 2026/Vendors/{Vendor Name} if missing. Upload each file, preserve the original filename, and return a table with vendor, destination URL, action taken, and any failures. Do not delete or overwrite files unless you ask me first.”
This works better because it defines the destination, safety behavior, folder convention, and expected verification output.
Recommended workflow for first runs
Start read-only. Ask the agent to confirm connections, identify the site, list subsites or libraries, and preview intended changes. Then run a small test with one file or one list item. For production use, require a final summary with object names, URLs, IDs, skipped items, and permission errors. This is especially important when using SharePoint Automation for Document Filing, because small path mistakes can create duplicate folders or place files in the wrong library.
SharePoint Automation skill FAQ
Is SharePoint Automation only for developers?
No. The skill is designed for agent-driven operations, so non-developers can use it if they understand their SharePoint structure. However, administrators or power users will have the easiest setup because they can authorize Microsoft 365 access, recognize site IDs, and resolve permission issues.
How is this better than an ordinary SharePoint prompt?
A normal prompt can explain SharePoint steps, but it cannot reliably execute them. The SharePoint Automation skill gives the agent tool-aware instructions for using Rube MCP with SharePoint REST and Microsoft Graph / OneDrive-backed actions. That reduces guesswork around which tools to call, what identifiers are required, and how to sequence discovery before modification.
What are the main boundaries?
The skill does not bypass Microsoft permissions, retention policies, sensitivity labels, or tenant restrictions. It also depends on the available Rube toolkit actions and the authenticated account’s access. If your workflow requires unsupported personal Microsoft accounts, heavily customized approval flows, or tenant-level administration outside the exposed tools, this skill may not be enough by itself.
When should I not install it?
Do not install SharePoint Automation if you only need one-time written instructions, cannot connect an MCP server, or do not have an organizational Microsoft 365 account. It is also a poor fit when your team cannot define destination sites, libraries, metadata requirements, or safe update rules; in that case, fix the information architecture first.
How to Improve SharePoint Automation skill
Improve SharePoint Automation results with better context
The most useful input is a compact operating brief: tenant or hostname, target site, library or list, folder policy, naming convention, metadata fields, overwrite rules, and required output format. Add examples of valid folder paths and list columns. If the agent must search first, specify whether it should match by filename, title, modified date, author, content, or metadata.
Prevent common failure modes
Common failures include missing active connections, personal Microsoft account authentication, wrong site_id format, insufficient permissions, ambiguous library names, and unsafe overwrite assumptions. Reduce risk by asking the agent to run discovery first, show the planned operation, and wait for approval before writes. For filing workflows, require a post-action verification step that retrieves the created or updated item.
Iterate after the first output
After the first run, improve the prompt with what actually happened: “The library is Shared Documents, not Documents,” “Use fiscal year folders,” or “Skip files already present with the same name and size.” Turn repeated corrections into reusable instructions. If the task involves lists, include exact internal column names where possible, because display names and internal field names may differ.
Repository reading path for maintainers
Start with SKILL.md; it contains the setup notes, tool dependencies, and core workflows. Because the current package has no extra scripts or reference files, improvements should go directly into the skill instructions: clearer discovery flow, safer write-operation rules, more examples for list and document handling, and troubleshooting notes for Microsoft 365 organizational account requirements.