AI agents call browse_catalog to retrieve information from Skill Hub without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves and queries information from a skill catalog. It enables discovery and filtering of existing skills but does not create, modify, delete, or execute anything. The filtering and sorting are applied to display options, not actions that change state. This is a classic Read category tool with minimal risk if misused by an AI agent.
From the tool's definition Tool description states 'Browse the skill catalog' and 'discovering skills' — both read-only operations. No mention of modification, deletion, execution, or financial operations.
Attacks that exploit this kind of access
Browse the skill catalog with filtering and sorting options. Good for discovering skills by category or popularity. It is categorised as a Read tool in the Skill Hub MCP Server, which means it retrieves data without modifying state.
Register the Skill Hub MCP server in PolicyLayer and add a rule for browse_catalog: 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 Skill Hub. Nothing to install.
browse_catalog 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 browse_catalog 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 browse_catalog. 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.
browse_catalog is provided by the Skill Hub MCP server (skillhub-club/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 →