AI agents use ssh_upload to create or update resources in Mcpx — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Mcpx environment.
ssh_upload transfers files to a remote system via SSH, which creates or modifies data on the destination. This is a Write operation (reversible via deletion). Severity is high because unauthorized file uploads could inject malicious code, overwrite critical configurations, or compromise remote systems.
From the tool's definition Tool name 'ssh_upload' combined with server description stating it enables 'file transfer' and server context showing shell/SSH capabilities. Empty tool description limits specificity but 'upload' is unambiguous write operation semantics.
Attacks that exploit this kind of access
ssh_upload. It is categorised as a Write tool in the Mcpx MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Mcpx MCP server in PolicyLayer and add a rule for ssh_upload: 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 Mcpx. Nothing to install.
ssh_upload 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 ssh_upload 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 ssh_upload. 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.
ssh_upload is provided by the Mcpx MCP server (rmednitzer/relay-shell). 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 →