share_point-automation
by ComposioHQshare_point-automation helps agents automate SharePoint sites, document libraries, lists, and content tasks through Rube MCP with live tool discovery before execution.
Score: 70/100. This is an acceptable but limited listing candidate: directory users can understand when to use it and how to start SharePoint automation through Rube MCP, but should expect a lightweight integration guide rather than a fully packaged workflow with scripts, reference files, or stable schemas.
- Clear purpose and trigger surface: automating SharePoint document libraries, sites, lists, and content management via Rube MCP.
- Includes concrete prerequisites and setup steps, including adding `https://rube.app/mcp`, checking `RUBE_SEARCH_TOOLS`, and activating the `share_point` connection with `RUBE_MANAGE_CONNECTIONS`.
- Strong operational guardrail to search tools first for current schemas, which is appropriate for a changing MCP/toolkit integration.
- Execution depends on an external Rube MCP server and an ACTIVE Composio SharePoint connection; there is no standalone script or local automation included.
- The skill relies heavily on dynamic `RUBE_SEARCH_TOOLS` discovery rather than providing fixed SharePoint tool schemas or detailed end-to-end examples, so agents may still need some guesswork.
Overview of share_point-automation skill
What share_point-automation does
share_point-automation is a Claude skill for automating Microsoft SharePoint work through Composio’s Rube MCP toolkit. It is designed for agents that need to discover SharePoint sites, document libraries, lists, and content-management actions, then call the correct Rube tools with current schemas instead of guessing tool names or request fields.
Best-fit users and workflows
This skill is a good fit for teams using SharePoint as an operational system of record: document library maintenance, site and list inspection, file organization, metadata updates, and repeatable admin or content workflows. It is most useful when you want an AI agent to help execute SharePoint actions safely through MCP rather than only draft instructions for a human.
Key differentiator: tool discovery first
The most important behavior in the share_point-automation skill is its requirement to call RUBE_SEARCH_TOOLS before execution. SharePoint tool schemas can vary by toolkit version and available connection context, so the skill prioritizes live discovery of tool slugs, input schemas, recommended execution steps, and known pitfalls before any workflow runs.
Adoption considerations
You need Rube MCP available in your client and an active SharePoint connection through RUBE_MANAGE_CONNECTIONS using the share_point toolkit. The repository path is composio-skills/share_point-automation, and the main file to inspect is SKILL.md; there are no extra scripts or reference folders, so most operational guidance lives directly in that file.
How to Use share_point-automation skill
share_point-automation install and setup context
Install the skill in a compatible skills environment, for example:
npx skills add ComposioHQ/awesome-claude-skills --skill share_point-automation
Then configure Rube MCP by adding https://rube.app/mcp as an MCP server in your client. The skill expects RUBE_SEARCH_TOOLS to be available. Use RUBE_MANAGE_CONNECTIONS with toolkit share_point; if the connection is not ACTIVE, complete the returned authentication flow before asking the agent to modify or retrieve SharePoint content.
Inputs the skill needs from you
For reliable share_point-automation usage, give the agent enough SharePoint context to narrow discovery and execution:
- Target object: site, document library, list, folder, file, or content item
- Desired action: search, list, create, update, upload, move, delete, or inspect
- Identifiers you already know: site name, URL, drive/library name, list name, item ID, file path, or folder path
- Safety rules: read-only first, require confirmation before writes, avoid deletes, preserve metadata
- Output format: summary, table, CSV-like rows, audit trail, or step-by-step execution log
A weak prompt is: “Clean up our SharePoint files.”
A stronger prompt is: “Using share_point-automation, first discover available SharePoint tools with RUBE_SEARCH_TOOLS. Then inspect the Marketing site’s ‘Campaign Assets’ document library, list folders modified in the last 90 days, and propose an archive plan. Do not move or delete anything without confirmation.”
Recommended workflow for agents
Start every run by confirming Rube MCP availability and SharePoint connection status. Next, call:
RUBE_SEARCH_TOOLS: queries=[{"use_case":"document libraries, sites, lists, and content management","known_fields":""}]
Use the returned schemas to choose the exact tool calls. For read/write workflows, ask the agent to perform a read-only discovery pass first, summarize the planned operations, then execute only after you approve. This pattern is especially important for document libraries and lists where names can be ambiguous and permissions may differ across sites.
Repository files to read first
Read SKILL.md before installing or modifying the skill. Focus on the sections titled Prerequisites, Setup, Tool Discovery, and Core Workflows. Because the skill has no companion scripts, rules, or resource files, its practical value depends on following the live tool-discovery instruction and not treating the examples as fixed schemas.
share_point-automation skill FAQ
Is share_point-automation only for SharePoint admins?
No. Admins can use it for site, list, and library operations, but content managers, operations teams, and workflow owners can also benefit if they have the right SharePoint permissions. The agent can only act through the connected account and available Rube tools.
How is this better than a normal SharePoint prompt?
A generic prompt can describe SharePoint steps, but it may invent APIs or outdated fields. The share_point-automation skill tells the agent to discover real Composio/Rube SharePoint tools first, retrieve current schemas, and build execution around the active connection. That reduces guesswork when moving from advice to action.
When should I not use this skill?
Do not use it when you only need general SharePoint training, UI navigation help, or a policy document. Also avoid write operations if you cannot verify the target site, library, list, or item identifiers. For high-risk actions such as deletes, permission changes, or bulk moves, require a read-only preview and explicit approval.
Does it fit Workflow Automation projects?
Yes, share_point-automation for Workflow Automation is a natural fit when SharePoint is one step in a larger process, such as collecting documents, updating list items, organizing submitted files, or preparing audit summaries. It works best when paired with clear triggers, target locations, and confirmation rules.
How to Improve share_point-automation skill
Improve share_point-automation prompts with concrete context
Better results come from replacing broad goals with operational detail. Include the SharePoint area, the entity type, the intended action, constraints, and what should happen after discovery. For example: “Find all files in the Finance site’s ‘Vendor Contracts’ library missing a renewal date metadata field; return a table with file name, path, last modified date, and proposed next action. Do not update metadata yet.”
Reduce common failure modes
The most common problems are stale assumptions about tool schemas, ambiguous site or library names, missing authentication, and unsafe bulk actions. Counter these by requiring RUBE_SEARCH_TOOLS first, checking RUBE_MANAGE_CONNECTIONS, using exact identifiers where possible, and separating planning from execution. If a tool call fails, ask the agent to re-check the discovered schema rather than retrying the same payload.
Add guardrails for higher-risk operations
For production SharePoint environments, define guardrails in the prompt: read-only discovery first, maximum item counts, dry-run summaries, confirmation before writes, and no deletes unless explicitly approved. Ask for an operation log that records target objects, selected tool slugs, input fields used, and results. This makes the skill easier to audit and safer for repeated workflows.
Iterate after the first output
After the first run, refine by adding missing identifiers, narrowing date ranges, excluding folders, or changing the output format. If the agent returns too much data, ask it to filter at the SharePoint query/tool level where supported instead of post-processing everything. If it returns too little, provide known site URLs, list names, library names, or sample file paths so the next share_point-automation guide run can target the right scope.