AI agents call search_companies to retrieve information from Zefix without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool searches company data in a public Swiss register. Search operations retrieve information without side effects or modifications. The sibling tools confirm a read-only pattern. No evidence suggests the tool creates, modifies, deletes, or executes actions. This is a standard Read category tool with low blast radius if misused by an agent—at worst it queries public company data.
From the tool's definition Tool name 'search_companies' indicates a query/search operation against the Zefix register. Server description specifies 'search and retrieve detailed information' and lists sibling tools like 'get_company_by_chid', 'get_company_by_uid',…
Attacks that exploit this kind of access
search_companies. It is categorised as a Read tool in the Zefix MCP Server, which means it retrieves data without modifying state.
Register the Zefix MCP server in PolicyLayer and add a rule for search_companies: 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 Zefix. Nothing to install.
search_companies 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 search_companies 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 search_companies. 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.
search_companies is provided by the Zefix MCP server (johnphilipp/mcp-server-zefix). 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 →