Crypt::Password CVE-2026-16235

Exploitation requires that the deployment actually uses Crypt::Password ≤0.28 to generate salts (the vulnerable rand()-based path in lib/Crypt/Password.pm) AND that the attacker already possesses the resulting password hashes, since predictable salts only provide advantage during offline cracking of an obtained hash store. The primary limiting factor is this prerequisite access to the hashes - the flaw does not itself grant remote access or code execution despite the published AV:N/PR:N vector. Impact is further bounded by whether the application relies on this module's salt rather than supplying its own CSPRNG-generated salt.

The signals conflict sharply and must be weighed against the actual mechanism. The published CVSS 3.1 vector (AV:N/AC:L/PR:N/UI:N, C:H/I:H/A:H = 9.8) implies trivial remote total compromise, but the described flaw is a library weakness in salt generation, not a directly network-reachable code-execution or disclosure path - the CVSS score appears substantially inflated relative to the description. Supporting a lower real-world priority: EPSS is only 0.14% (4th percentile), there is no KEV listing, no known POC, and SSVC reports Exploitation: none. SSVC does flag Automatable: yes and Technical Impact: total, reflecting that once relevant, weak salts affect the whole hash corpus mechanically. Net assessment: this is a genuine cryptographic-hygiene defect but a low-urgency, defense-in-depth issue for most deployments, not a 9.8-grade emergency - exploitation is contingent on an attacker already holding the password hashes. Verify severity with the vendor/CPANSec advisory rather than acting on the raw 9.8.

An attacker who exfiltrates a password-hash store (via SQL injection, backup theft, or insider access) from an application using Crypt::Password ≤0.28 can predict or reconstruct the weakly generated salts, then run precomputation or optimized offline cracking to recover plaintext passwords far faster than properly salted hashes would allow. No public POC is known, and the attack presupposes prior access to the hashed credentials rather than direct remote exploitation of the library.