AI agents call search_exercises to retrieve information from Hevy MCP without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
Search operations typically retrieve data without modifying state. The tool fits the pattern of Read category tools (search, list, get, fetch) evident in the server's sibling tools. While the empty description reduces certainty slightly, the name and server context indicate this is a retrieval function with no side effects. Severity is low because querying exercise data carries minimal risk of harm if misused.
From the tool's definition Tool name 'search_exercises' indicates a query/search operation. Server description states the system enables 'reading workout history' and sibling tools include query operations like 'get_routine', 'get_workout', 'list_routines', 'list_workouts'.
Attacks that exploit this kind of access
search_exercises. It is categorised as a Read tool in the Hevy MCP MCP Server, which means it retrieves data without modifying state.
Register the Hevy MCP server in PolicyLayer and add a rule for search_exercises: 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 Hevy MCP. Nothing to install.
search_exercises 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_exercises 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_exercises. 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_exercises is provided by the Hevy MCP server (zachsai/hevy-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 →