🛡️ CVE-2026-54894 — guardian
Description
Atom-table exhaustion denial of service in Guardian via unbounded atom creation from binary keys
Summary
Allocation of Resources Without Limits or Throttling in ueberauth guardian allows denial of service via unbounded atom creation from attacker-influenced binary input.
Guardian.Plug.Keys derives connection and session namespace keys by passing arbitrary binaries to String.to\_atom/1. base\_key/1 in lib/guardian/plug/keys.ex converts any binary into the atom :"guardian\_<input>", and the derived helpers claims\_key/1, resource\_key/1, and token\_key/1 create a second atom on top of that. key\_from\_other/1 likewise converts a regex-captured binary through String.to\_atom/1. The public specs advertise String.t() as a valid argument, so passing a string is documented usage, and higher-level entry points such as Guardian.Plug.current\_token(conn, key: key) thread the caller-supplied key straight into these functions.
String.to\_atom/1 creates a brand-new atom for every previously unseen binary, atoms are never garbage collected, and the BEAM atom table is fixed at roughly 1,048,576 entries by default. An application that routes attacker-influenced data (a tenant identifier, header, or other request input) into a Guardian key therefore mints one permanent atom per distinct value. A modest stream of varied, unauthenticated input permanently consumes the atom table and crashes the BEAM node, taking down every application running on it.
This issue affects guardian: from 0.1.0 before 2.4.1.
Workaround
Do not derive Guardian keys from untrusted input. Use a fixed, hardcoded set of namespace keys, or validate the value against a bounded allowlist of known keys, before passing it as the :key option.
Configuration
Only applications that derive a Guardian key (the conn/session namespace) from attacker-influenced data are exploitable. This is the case when a caller-supplied value such as a tenant identifier, request header, or other request input is passed as the :key option to entry points like Guardian.Plug.current\_token/2. Applications that use a static namespace (the default :default key or hardcoded atoms) are not affected.
How this vulnerability can be exploited
This issue can be reached with local access to the system, attack complexity is low, an attacker needs no privileges on the target. No user interaction is required. Rated impact: confidentiality none, integrity none, availability high.
Weakness class
CVE-2026-54894 is classified as CWE-770: Allocation of Resources Without Limits. Resources are allocated on request with no cap, so a client can exhaust them.
Affected software
CVE-2026-54894 is recorded against 1 package.
- guardian
Timeline and source
Published on 1 August 2026 and last revised on 6 August 2026. A public exploit is known to exist, which raises the urgency of patching considerably. A vendor advisory or fix has been published. Record sourced from OSV.
References
github.com (Advisory)
cna.erlef.org (Web)
github.com (Fix)
hex.pm (Package)
Details
CVSS:4.0/AV:L/AC:L/AT:P/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:H
Affected Packages
| Software | From version | Fixed in |
|---|---|---|
| guardian | — | — |
References
Similar Threats
- High CVE-2025-3718
- High CVE-2025-3719
- Medium CVE-2025-40885
- High CVE-2025-40886
- Medium CVE-2024-4465
Exploit Protection
Are you running guardian?
CVE-2026-54894 carries CVSS 7.5 High rating and a public exploit already exists. BotEraser checks your installation against this and other known CVE records, and blocks IPs associated with exploit activity.
Check My Site For CVE-2026-54894 →No credit card required · Results in minutes
ⓘ Data Notice: The information presented above has been compiled from publicly available internet sources. Boteraser aggregates this data solely for informational purposes and does not independently classify, evaluate, or endorse any findings about the vulnerabilities listed. The accuracy and completeness of this information is the sole responsibility of the original publishers. Boteraser and its operators accept no liability for any decisions made based on this data.