Upload a file for use by assistants
AI agents use upload-file to create or update resources in OpenAI Assistant MCP Server — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your OpenAI Assistant MCP Server environment.
This tool creates or stores data (a file upload) in OpenAI's infrastructure, which is a Write operation. It is not Destructive because the file can be deleted. It is not Execute because it does not run code or trigger external operations based on file content.
From the tool's definition Tool name 'upload-file' and description 'Upload a file for use by assistants' indicates creation of new file data in OpenAI's system. The operation is reversible via the sibling tool 'delete-file'.
Attacks that exploit this kind of access
Upload a file for use by assistants. It is categorised as a Write tool in the OpenAI Assistant MCP Server MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the OpenAI Assistant MCP Server MCP server in PolicyLayer and add a rule for upload-file: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches OpenAI Assistant MCP Server. Nothing to install.
upload-file is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the upload-file rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for upload-file. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
upload-file is provided by the OpenAI Assistant MCP Server MCP server (snilld-ai/openai-assistant-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →