spec_get
AI agents call spec_get to retrieve information from Openapi Tools without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool is part of an OpenAPI inspection server that retrieves specification details from local files. 'Get' operations are fundamentally read-only with no side effects. The tool fits the Read category: retrieves or queries data without modification. Severity is low because inspecting OpenAPI specs poses minimal risk even if misused by an AI agent.
From the tool's definition Tool name 'spec_get' combined with server description stating it 'retrieves spec details' indicates a read operation. The server's purpose is 'inspecting local OpenAPI specification files by listing and retrieving spec details'.
Attacks that exploit this kind of access
spec_get. It is categorised as a Read tool in the Openapi Tools MCP Server, which means it retrieves data without modifying state.
Register the Openapi Tools MCP server in PolicyLayer and add a rule for spec_get: 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 Openapi Tools. Nothing to install.
spec_get 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 spec_get 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 spec_get. 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.
spec_get is provided by the Openapi Tools MCP server (kardaj/openapi-tools-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 →