Search roles by name, description, or tags
AI agents call roles_search to retrieve information from Claude Role Library without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves and filters role information from the local YAML-based storage without creating, modifying, deleting, or executing any operations. It has no side effects and poses minimal security risk—the worst case is that an AI agent discovers role definitions it could then apply, which is the intended functionality of this system.
From the tool's definition Tool name 'roles_search' combined with description 'Search roles by name, description, or tags' indicates a query operation that retrieves data without modification.
Attacks that exploit this kind of access
Search roles by name, description, or tags. It is categorised as a Read tool in the Claude Role Library MCP Server, which means it retrieves data without modifying state.
Register the Claude Role Library MCP server in PolicyLayer and add a rule for roles_search: 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 Claude Role Library. Nothing to install.
roles_search 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 roles_search 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 roles_search. 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.
roles_search is provided by the Claude Role Library MCP server (tony427/claude-role-library). 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 →