Auditar vulnerabilidades de seguridad
AI agents call npm_audit to retrieve information from NPM MCP Server without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
npm_audit performs vulnerability scanning and reporting only. It queries npm's vulnerability database and reports findings without installing packages, executing code, or making any modifications to the project. This is a classic Read operation: it retrieves data about potential security issues but has no side effects on the system or codebase.
From the tool's definition Tool name 'npm_audit' and description 'Auditar vulnerabilidades de seguridad' (Audit security vulnerabilities) indicates a read-only security scanning operation that retrieves vulnerability information without modifying or executing code.
Attacks that exploit this kind of access
Auditar vulnerabilidades de seguridad. It is categorised as a Read tool in the NPM MCP Server MCP Server, which means it retrieves data without modifying state.
Register the NPM MCP Server MCP server in PolicyLayer and add a rule for npm_audit: 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 NPM MCP Server. Nothing to install.
npm_audit 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 npm_audit 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 npm_audit. 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.
npm_audit is provided by the NPM MCP Server MCP server (kpangaa/npm-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 →