AI agents call skills_get_reference to retrieve information from Skills without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool appears designed to fetch or retrieve reference information from a skills registry. No verbs suggesting modification (create, update, delete, execute) are present in the name. The empty description lowers confidence slightly, but the naming pattern and context of a registry with find/get/list operations point to read-only access.
From the tool's definition Tool name 'skills_get_reference' combined with sibling tools like 'skills_get_asset', 'skills_get_body', and 'skills_get_options' suggests a retrieval operation.
Attacks that exploit this kind of access
skills_get_reference. It is categorised as a Read tool in the Skills MCP Server, which means it retrieves data without modifying state.
Register the Skills MCP server in PolicyLayer and add a rule for skills_get_reference: 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 Skills. Nothing to install.
skills_get_reference 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 skills_get_reference 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 skills_get_reference. 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.
skills_get_reference is provided by the Skills MCP server (jignesh-ponamwar/skills-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 →