AI agents call extract-dicom-metadata to retrieve information from DICOM-MCP 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 that retrieves metadata from medical images without altering, creating, or destroying data. However, severity is elevated to 'medium' rather than 'low' because DICOM metadata in healthcare contexts can contain Protected Health Information (PHI) including patient names, medical record numbers, study dates, and other sensitive personal/medical details.
From the tool's definition Tool description states 'Extract detailed metadata from a DICOM file' — the verb 'extract' indicates retrieval of information with no modification. The tool reads and queries DICOM file metadata.
Attacks that exploit this kind of access
Extract detailed metadata from a DICOM file. It is categorised as a Read tool in the DICOM-MCP MCP Server, which means it retrieves data without modifying state.
Register the DICOM- MCP server in PolicyLayer and add a rule for extract-dicom-metadata: 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 DICOM-MCP. Nothing to install.
extract-dicom-metadata 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 extract-dicom-metadata 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 extract-dicom-metadata. 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.
extract-dicom-metadata is provided by the DICOM- MCP server (shaunporwal/dicom-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 →