search_source
AI agents call search_source to retrieve information from Unreal Source without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves or queries data from the Unreal Engine source code index without modifying, deleting, or executing anything. It performs a search operation across source files, consistent with other analysis tools on the server. The lack of a description is a minor confidence reduction, but the context strongly indicates this is a search/query operation with no side effects.
From the tool's definition Tool name 'search_source' combined with server context showing 'full-text search across all engine source files' and sibling tools like 'find_callees', 'find_references', 'read_file' which are all Read operations.
Attacks that exploit this kind of access
search_source. It is categorised as a Read tool in the Unreal Source MCP Server, which means it retrieves data without modifying state.
Register the Unreal Source MCP server in PolicyLayer and add a rule for search_source: 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 Unreal Source. Nothing to install.
search_source 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_source 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_source. 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_source is provided by the Unreal Source MCP server (tumourlove/deprecated-unreal-source-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 →