AI agents call get_tokens to retrieve information from Musea without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves design token data (colors, spacing, typography) from configuration files without modifying, executing, or deleting anything. It is a straightforward read operation with minimal blast radius—worst case, an agent might waste resources reading large token files, but no system state is altered.
From the tool's definition Tool name is 'get_tokens' and description explicitly states 'Read design tokens...from a Style Dictionary-compatible JSON file or directory.' The verb 'Read' combined with 'get' indicates a retrieval operation with no side effects.
Attacks that exploit this kind of access
Read design tokens (colors, spacing, typography, etc.) from a Style Dictionary-compatible JSON file or directory. Auto-detects common paths if not specified. It is categorised as a Read tool in the Musea MCP Server, which means it retrieves data without modifying state.
Register the Musea MCP server in PolicyLayer and add a rule for get_tokens: 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 Musea. Nothing to install.
get_tokens 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 get_tokens 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 get_tokens. 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.
get_tokens is provided by the Musea MCP server (@vizejs/musea-mcp-server). 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 →