snowball_search
AI agents call snowball_search to retrieve information from Paper Search without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Snowball searching is a standard research methodology for tracing citations and identifying related papers—a read-only operation that retrieves data without modifying or executing anything. The server's explicit mention of 'citation network tracing' as a feature further supports this classification. No data is created, deleted, or executed; the tool queries and retrieves information only.
From the tool's definition The tool name 'snowball_search' combined with the server's documented functionality (searching and downloading academic papers, citation network tracing) suggests a search operation.
Attacks that exploit this kind of access
snowball_search. It is categorised as a Read tool in the Paper Search MCP Server, which means it retrieves data without modifying state.
Register the Paper Search MCP server in PolicyLayer and add a rule for snowball_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 Paper Search. Nothing to install.
snowball_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 snowball_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 snowball_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.
snowball_search is provided by the Paper Search MCP server (its-antony/paper-search). 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 →