AI agents call string_result_tool to retrieve information from Study without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool appears to simply return a string result, suggesting a read/passthrough operation with no side effects. However, the description is minimal and uninformative, so confidence is low. Given the server context (math/demo tools), this is likely a utility tool that formats or returns a string. No evidence of write, execute, destructive, or financial behavior.
From the tool's definition 'For string results, return the string directly' — purely returns a string value with no described side effects
Attacks that exploit this kind of access
For string results, return the string directly. It is categorised as a Read tool in the Study MCP Server, which means it retrieves data without modifying state.
Register the Study MCP server in PolicyLayer and add a rule for string_result_tool: 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 Study. Nothing to install.
string_result_tool 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 string_result_tool 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 string_result_tool. 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.
string_result_tool is provided by the Study MCP server (lucs1590/study-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 →