Test tool that returns a secret word
AI agents call secret_phrase to retrieve information from MCP Server Example without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
The tool appears to simply return a static secret phrase/word, which is a read-only operation with no side effects. Severity is low as it likely returns a hardcoded test value, though there is mild concern that it could expose sensitive information. Confidence is moderate because the description is vague ('Test tool that returns a secret word') and does not clarify what the 'secret' entails or where it comes from.
From the tool's definition 'returns a secret word' — the tool retrieves and returns a fixed value
Attacks that exploit this kind of access
Test tool that returns a secret word. It is categorised as a Read tool in the MCP Server Example MCP Server, which means it retrieves data without modifying state.
Register the MCP Server Example MCP server in PolicyLayer and add a rule for secret_phrase: 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 MCP Server Example. Nothing to install.
secret_phrase 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 secret_phrase 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 secret_phrase. 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.
secret_phrase is provided by the MCP Server Example MCP server (sriharsha-gvkss/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 →