Jawn: Uncontrolled nesting depth in JSON parser
GHSA-cc4v-rvgp-2pf3 · CVE-2026-59990
Published · Modified
AI SAST
Find this class of vulnerability in your own code
Corgea's AI-native static analysis detects vulnerabilities like this one across your repositories, ranks them by exploitability, and returns review-ready fixes.
Description
The Jawn parser before 1.6.1 is vulnerable to a denial of service attack via untrusted input.
Impact
A remote attacker who can submit JSON to any jawn-backed parse method can exhaust JVM heap and trigger java.lang.OutOfMemoryError. This is treated by Scala as a fatal error and not typically handled by scala.util.Try or cats.effect.IO.
Patches
Version 1.6.1 introduces a configurable nesting-depth limit (Parser#maxDepth, default 4096). Inputs deeper than the limit fail with a recoverable ParseException instead of exhausting heap.
Users who require deeper nesting may override maxDepth on a Parser subclass.
Workarounds
- Enforce an input size limit small enough that the resulting context stack cannot exhaust heap, (e.g. http4s
EntityLimiter). - Pre-scan untrusted input and reject documents whose maximum delimiter nesting exceeds a threshold before handing them to jawn.
References
- WEB https://github.com/typelevel/jawn/security/advisories/GHSA-cc4v-rvgp-2pf3
- WEB https://github.com/typelevel/jawn/commit/191cb3a44e77f1afab439ee636bf66bdf3c54a04
- WEB https://github.com/typelevel/jawn/commit/6219666641f9408498f85868f835e17bd8a72fed
- WEB https://github.com/typelevel/jawn/commit/93ac93e9c992c11b4c03d5455d8551f9fb24da1b
- WEB https://github.com/typelevel/jawn/commit/f6ace7e0db715de1a8c4618bed9378333a5c2214
- PACKAGE https://github.com/typelevel/jawn
- WEB https://github.com/typelevel/jawn/releases/tag/v1.7.0
Ready to move
Start Securing
Free, no credit card | First findings in minutes