AI agents call search-roles to retrieve information from Aria without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool performs a read-only search across the W3C WAI-ARIA specification database. It retrieves information based on keyword matching and returns results without creating, modifying, deleting, or executing any operations. The blast radius of misuse is minimal — the worst outcome would be retrieving irrelevant or all role information, with no harmful consequences to systems or data.
From the tool's definition Tool name 'search-roles' and description state it 'Search for ARIA roles by keyword in their name or description' — a pure query operation that retrieves and returns matching role information with no side effects or modifications.
Attacks that exploit this kind of access
Search for ARIA roles by keyword in their name or description. It is categorised as a Read tool in the Aria MCP Server, which means it retrieves data without modifying state.
Register the Aria MCP server in PolicyLayer and add a rule for search-roles: 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 Aria. Nothing to install.
search-roles 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-roles 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-roles. 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-roles is provided by the Aria MCP server (joe-watkins/aria-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 →