CVE-2026-48147 is a medium-severity cross-site request forgery (CSRF) vulnerability in @budibase/backend-core (npm), affecting versions < 3.35.4. It is fixed in 3.35.4.
Summary The buildMatcherRegex() / matches() functions in packages/backend-core/src/middleware/matchers.ts share the same structural root cause as the recently patched CVE-2026-31816: route patterns are compiled into unanchored regular expressions and tested against ctx.request.url, which includes the full query string. The CSRF middleware in the Budibase Worker uses this matching system to decide whether to skip CSRF token validation. An unauthenticated attacker can forge state-changing cross-origin requests against any Worker API endpoint by injecting a public route pattern into the query string, causing the CSRF middleware to skip token validation entirely. This allows actions such as sending admin invites, modifying global configuration, and managing users without a valid CSRF token. CVE-2026-31816 fixed the same unanchored-regex-on-full-URL bug in server/middleware/utils.ts but left backend-core/middleware/matchers.ts untouched. Details Root cause, packages/backend-core/src/middleware/matchers.ts: Two compounding bugs identical to the patched CVE: new RegExp(route), no ^ start anchor, no $ end anchor. ctx.request.url, full URL string including ?query=value, not just the path. CSRF middleware, packages/backend-core/src/middleware/csrf.ts: Worker registration, packages/worker/src/api/index.ts: buildMatcherRegex compiles "/api/global/auth/:tenantId" into the regex /api/global/auth/. (via PARAMREGEX replacing /:tenantId → /.). Since the regex is unanchored, it matches the substring "/api/global/auth/" anywhere in ctx.request.url, including inside a query string parameter on a completely different endpoint. Triggering condition: ctx.request.url = "/api/global/users/invite?x=/api/global/auth/evil" new RegExp("/api/global/auth/.").test(ctx.request.url) → true (substring found in query string) ctx.request.method === "POST" → true matches() returns the pattern entry → CSRF skipped The protected user-invite POST proceeds without any CSRF token Additional affected middleware (same matches() call): | Middleware | Pattern list | Security effect when bypassed | |---|---|---| | csrf() | NOCSRFENDPOINTS | CSRF token validation skipped | | tenancy() | NOTENANCYENDPOINTS | allowNoTenant = true, bypasses tenant ID requirement | | authenticated() | PUBLICENDPOINTS | Marks endpoint as publicEndpoint = true | The NOTENANCYENDPOINTS entry { route: "/api/system", method: "ALL" } compiles to /api/system (no param replacement). Since method: "ALL" matches every HTTP verb and the pattern is unanchored, any request with ?x=/api/system/x in its URL matches, potentially bypassing tenant isolation in multi-tenant deployments. PoC Prerequisites: Victim user is logged into a Budibase Worker instance (e.g., https://budibase.target.com) in their browser. Attacker hosts a page at https://evil.com. Step 1, Verify CSRF is normally enforced: Step 2, CSRF bypass via query string injection: Step 3, Cross-site exploitation (victim visits attacker page): When the authenticated victim visits https://evil.com/csrf.html, the browser sends the cross-origin POST including the victim's session cookie. The Worker's CSRF middleware matches /api/global/auth/. in the query string, skips validation, and processes the admin invite. Impact An unauthenticated attacker can forge state-changing requests on behalf of any authenticated Budibase admin by injecting a public endpoint pattern into the query string of a Worker API call. Operations exposed by the Worker that become exploitable via CSRF include: User management: send admin/operator invites, delete users, modify roles Global configuration: update authentication settings (SSO, OIDC, SMTP), change branding Tenant administration: in multi-tenant instances, tenant isolation bypass via NOTENANCYENDPOINTS (/api/system match) combined with the CSRF bypass allows cross-tenant administrative actions The vulnerability affects all Budibase self-hosted deployments with an internet-accessible Worker service up to and including version 3.32.3, which is the latest released version. It is a direct sibling of CVE-2026-31816 and stems from the same unanchored-regex-on-full-URL pattern in packages/backend-core/src/middleware/matchers.ts, which was not addressed by the patch for that CVE.
A victim's authenticated browser session is used to submit forged requests to an application that cannot distinguish them from legitimate ones. Typical impact: state-changing actions performed as the victim without their consent.
CVE-2026-48147 has a CVSS score of 6.5 (Medium). The vector is network-reachable, no privileges required, and user interaction required. A CVSS score reflects the worst-case severity of the vulnerability, not your specific exposure. Whether this affects your application depends on whether the vulnerable code is present and reachable in your environment.
A fixed version is available (3.35.4). Upgrading removes the vulnerable code path.
npm
@budibase/backend-core (< 3.35.4)@budibase/backend-core → 3.35.4 (npm)Severity tells you how bad this could be in the worst case. It does not tell you whether you are exposed. Exploitability and impact are functions of runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A vulnerable package can sit in your dependency tree and never run.
Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter instead of chasing every advisory.
Kodem's runtime-powered SCA identifies whether CVE-2026-48147 is reachable in your applications. Explore open-source security for your team.
See if CVE-2026-48147 is reachable in your applications. Get a demo
Already deployed Kodem? See CVE-2026-48147 in your environment →Upgrade @budibase/backend-core to 3.35.4 or later to resolve this vulnerability.
Kodem Kai can prioritize this vulnerability in your dependency tree and generate a fix recommendation.
CVE-2026-48147 is a medium-severity cross-site request forgery (CSRF) vulnerability in @budibase/backend-core (npm), affecting versions < 3.35.4. It is fixed in 3.35.4. A victim's authenticated browser session is used to submit forged requests to an application that cannot distinguish them from legitimate ones.
CVE-2026-48147 has a CVSS score of 6.5 (Medium). This score reflects the worst-case severity of the vulnerability, not your specific exposure. Whether it represents real risk in your environment depends on whether the vulnerable code is present and reachable.
@budibase/backend-core (npm) versions < 3.35.4 is affected.
Yes. CVE-2026-48147 is fixed in 3.35.4. Upgrade to this version or later.
Whether CVE-2026-48147 is exploitable in your environment depends on whether the vulnerable code is present and reachable. A CVSS score is a worst-case rating; it does not account for your specific deployment, configuration, or usage patterns. Kodem, an Intelligent Application Security platform, uses runtime intelligence to show which vulnerabilities actually execute in production, so you can focus on the ones that represent real risk. Get a demo
Exploitability and impact are not fixed properties of a CVE. They depend on runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A high CVSS score on a dependency that never runs is not the same as real risk. Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter.
Upgrade @budibase/backend-core to 3.35.4 or later.