AI agents call references to retrieve information from Symbols without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Even though references only reads data, uncontrolled read access leaks sensitive information and racks up API costs — an agent caught in a retry loop can make thousands of calls a minute without anyone noticing.
Attacks that exploit this kind of access
Find semantic references to the symbol at a file position across the workspace. It is categorised as a Read tool in the Symbols MCP Server, which means it retrieves data without modifying state.
Register the Symbols MCP server in PolicyLayer and add a rule for references: 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 Symbols. Nothing to install.
references 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 references 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 references. 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.
references is provided by the Symbols MCP server (@p1va/symbols). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.