AI agents call sg_search_datasets to retrieve information from Sgdata without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool performs a simple keyword search across public Singapore government datasets. It has no side effects, does not execute code or commands, does not modify data, and does not involve financial transactions. The blast radius of misuse is minimal - an agent could only retrieve information about available datasets, which is already public data on data.gov.sg. This is a straightforward Read operation.
From the tool's definition Tool description states 'Search data.gov.sg for datasets matching a free-text keyword query' - this is a read-only search operation that retrieves information without modifying, executing, or deleting anything.
Attacks that exploit this kind of access
Search data.gov.sg for datasets matching a free-text keyword query. It is categorised as a Read tool in the Sgdata MCP Server, which means it retrieves data without modifying state.
Register the Sgdata MCP server in PolicyLayer and add a rule for sg_search_datasets: 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 Sgdata. Nothing to install.
sg_search_datasets 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 sg_search_datasets 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 sg_search_datasets. 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.
sg_search_datasets is provided by the Sgdata MCP server (sypherin/sgdata-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 →