AI agents call get_recent_cves to retrieve information from Nvd Cve without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool queries and retrieves historical CVE vulnerability data with a configurable time window. It has no ability to modify data, execute code, delete records, or affect external systems. The blast radius of misuse is minimal—an agent could retrieve excessive CVE data but cannot cause harm beyond consuming API quota or bandwidth. This is a straightforward Read operation.
From the tool's definition Tool description states 'Fetch recent CVEs' and server description confirms it 'retrieve[s] CVE data from the NVD API'. The operation is a data retrieval with no modification, creation, deletion, or side effects.
Attacks that exploit this kind of access
Fetch recent CVEs from the last days_back days (max 120 due to NVD API limit). It is categorised as a Read tool in the Nvd Cve MCP Server, which means it retrieves data without modifying state.
Register the Nvd Cve MCP server in PolicyLayer and add a rule for get_recent_cves: 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 Nvd Cve. Nothing to install.
get_recent_cves 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_recent_cves 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_recent_cves. 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_recent_cves is provided by the Nvd Cve MCP server (millsks/nvd-cve-mcp-server). 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 →