AI agents call download_media to retrieve information from Whatsapp without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Downloading and returning media content is fundamentally a data retrieval operation. It does not create, modify, delete, execute code, or involve financial transactions. The read-only nature of the server reinforces this classification. The minimal risk is exposure of sensitive chat media if an agent inappropriately requests it, but the tool itself performs no side effects.
From the tool's definition Tool description states it 'Download a media item' and 'return its bytes inline' — a retrieval operation with no modification or deletion of data.
Attacks that exploit this kind of access
Download a media item (audio/image/video/document) and return its bytes inline as MCP content. It is categorised as a Read tool in the Whatsapp MCP Server, which means it retrieves data without modifying state.
Register the Whatsapp MCP server in PolicyLayer and add a rule for download_media: 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 Whatsapp. Nothing to install.
download_media is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the download_media 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 download_media. 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.
download_media is provided by the Whatsapp MCP server (joaohts/whatsapp-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 →