AI agents call get_sleep to retrieve information from Garmin without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool queries personal health metrics without modifying, executing external commands, or causing irreversible changes. While sleep data is personal/sensitive, the read-only nature and lack of blast radius from misuse places it in the Read category with low severity. The confidence is high given the explicit 'read-only' server description.
From the tool's definition Tool is described as 'read-only MCP tools' that retrieve sleep data (duration, stages, score, HRV). The name 'get_' and verb 'retrieve' pattern confirm data retrieval with no side effects.
Attacks that exploit this kind of access
Sleep duration, sleep stages, sleep score, and overnight HRV. It is categorised as a Read tool in the Garmin MCP Server, which means it retrieves data without modifying state.
Register the Garmin MCP server in PolicyLayer and add a rule for get_sleep: 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 Garmin. Nothing to install.
get_sleep 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 get_sleep 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 get_sleep. 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.
get_sleep is provided by the Garmin MCP server (tyler-irving/garmin-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 →