Skip to main content

Security

Security hooks are configured through LioranManagerOptions.security (advanced).

This is embedded authorization/auditing for the engine (not the HTTP server). It is useful when:

  • you run multi-tenant workloads inside one process, or
  • you want a custom authorization layer around collections and documents.

Enable + authorize hook​

import { LioranManager } from "@liorandb/core";

const manager = new LioranManager({
  rootPath: "./.liorandb",
  security: {
    enabled: true,
    authorize: async (ctx) => {
      // Decide whether to allow the action.
      // Returning/throwing semantics depend on your hook implementation.
      return ctx;
    },
  },
});

Auditing​

import { LioranManager } from "@liorandb/core";

const manager = new LioranManager({
  rootPath: "./.liorandb",
  security: {
    enabled: true,
    audit: {
      enabled: true,
      dir: "./.liorandb/audit",
      flushDelayMs: 500,
    },
  },
});

Principals​

When security is enabled, operations may be associated with a principal (the caller identity). How you set or propagate principals depends on your app’s structure.