PraisonAI versions prior to 1.5.115 contain a flaw in the MultiAgentLedger component that allows duplication of agent identifiers. This lack of enforcement enables an attacker to register agents with the same ID, thereby accessing shared ledger data and exposing sensitive system prompts and conversation histories. The vulnerability is classified as CWE‑668 and results in the unintended disclosure of confidential information to entities that can register duplicate agents.

The affected product is PraisonAI; any deployment running a version older than 1.5.115 is impacted. No specific sub‑components outside the MultiAgentLedger are identified as vulnerable.

The CVSS score of 7.1 indicates a high severity impact. EPSS data is not available, so the current exploitation probability cannot be quantified, and the vulnerability is not listed in CISA KEV. Based on the description, the attack vector is inferred to require the ability to register agents—likely through the public or internal API used for agent creation—however the exact environment prerequisites are not specified in the advisory.