Content-Security-Policy: trusted-types directive

Limited availability

This feature is not Baseline because it does not work in some of the most widely-used browsers.

The HTTP Content-Security-Policy (CSP) trusted-types directive is used to specify an allowlist of Trusted Type policy names that a website can create using trustedTypes.createPolicy().

This prevents website code from creating unexpected policies, making it easier to audit trusted type code (createPolicy() will throw an exception if it is passed a name which was not listed in trusted-types).

Note: The require-trusted-types-for directive must be set to enable enforcement of trusted types, and the trusted-types-eval keyword is used to relax restrictions on eval() and Function() when trusted types are enabled.

See Trusted Type API for more information.

Syntax

http

Content-Security-Policy: trusted-types;
Content-Security-Policy: trusted-types 'none';
Content-Security-Policy: trusted-types <policyName>;
Content-Security-Policy: trusted-types <policyName> <policyName> 'allow-duplicates';

<policyName>: A valid policy name consists only of alphanumeric characters, or one of -#=_/@.%. A star (*) as a policy name instructs the user agent to allow any unique policy name (allow-duplicates may relax that further).
'none': Disallows creating any Trusted Type policy (same as not specifying any <policyName>).
'allow-duplicates': Allows for creating policies with a name that was already used.

Examples

// Content-Security-Policy: trusted-types foo bar 'allow-duplicates';

if (typeof trustedTypes !== "undefined") {
  const policyFoo = trustedTypes.createPolicy("foo", {});
  const policyFoo2 = trustedTypes.createPolicy("foo", {});
  const policyBaz = trustedTypes.createPolicy("baz", {}); // Throws and dispatches a SecurityPolicyViolationEvent.
}