Search for observations
AI agents call search_observations to retrieve information from FHIR MCP Server without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves healthcare observations without modifying data. It is a Read operation. Severity is high because observations contain sensitive protected health information (PHI) that could expose patient medical histories, test results, and clinical findings if misused by an AI agent without proper access controls.
From the tool's definition Tool name 'search_observations' and description 'Search for observations' indicate data retrieval. The FHIR MCP Server context shows this is medical data access. Observations in FHIR are clinical measurements and findings (vital signs, lab results, etc.).
Attacks that exploit this kind of access
Search for observations. It is categorised as a Read tool in the FHIR MCP Server MCP Server, which means it retrieves data without modifying state.
Register the FHIR MCP Server MCP server in PolicyLayer and add a rule for search_observations: 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 FHIR MCP Server. Nothing to install.
search_observations 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_observations 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_observations. 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_observations is provided by the FHIR MCP Server MCP server (jcafazzo/fhir-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 →