visual_search
AI agents call visual_search to retrieve information from Pinterest Vision without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Despite the description being empty, the tool name 'visual_search' and its position among read-oriented sibling tools (search, analyze, ingest) strongly indicate it retrieves or queries visual data from Pinterest without modifying or deleting content. No evidence suggests side effects, execution of arbitrary code, financial transactions, or destructive operations.
From the tool's definition Tool is named 'visual_search' and appears among sibling tools like 'pinterest_search' and 'pinterest_analyze' on a server described as providing 'visual intelligence' and 'search Pinterest.' The naming pattern and context suggest this performs…
Attacks that exploit this kind of access
visual_search. It is categorised as a Read tool in the Pinterest Vision MCP Server, which means it retrieves data without modifying state.
Register the Pinterest Vision MCP server in PolicyLayer and add a rule for visual_search: 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 Pinterest Vision. Nothing to install.
visual_search 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 visual_search 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 visual_search. 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.
visual_search is provided by the Pinterest Vision MCP server (kreminskaya/pinterest-vision-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 →