Given a natural-language use case, recommend MCP servers filtered by trust. Example:
AI agents call recommend to retrieve information from Agentforge Trust without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a Read operation: it queries and returns filtered recommendations. No data is created, modified, deleted, or executed. Even though it evaluates against policies, the tool itself only retrieves and suggests information. The blast radius of misuse is minimal—an agent could request irrelevant recommendations but cannot cause damage via this tool alone.
From the tool's definition Tool performs a search/recommendation query based on trust filtering; returns suggestions without modifying data. Description shows it 'recommend[s] MCP servers filtered by trust' — a retrieval operation with no side effects.
Attacks that exploit this kind of access
Given a natural-language use case, recommend MCP servers filtered by trust. Example:. It is categorised as a Read tool in the Agentforge Trust MCP Server, which means it retrieves data without modifying state.
Register the Agentforge Trust MCP server in PolicyLayer and add a rule for recommend: 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 Agentforge Trust. Nothing to install.
recommend 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 recommend 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 recommend. 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.
recommend is provided by the Agentforge Trust MCP server (kovy/agentforge-trust-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 →