CVE-2026-56074

What it is

PraisonAI before 1.5.128 caches tool approval decisions by tool name only, not by invocation arguments, allowing subsequent execute_command calls to bypass approval prompts. Attackers can exploit this by obtaining initial approval for a benign command, then silently exfiltrate API keys and credentials via subsequent shell commands without user consent.

At a glance

Exploitability

Can be exploited locally on the machine without logging in, but needs a user to act. High risk of data exposure.

Exploit odds

0.2% chance of exploitation in the next 30 days (EPSS).

Fix

A patch, advisory or fixed version is referenced. Upgrade or apply it.

Timeline

1. 2026-06-18 23:16 UTC PublishedDisclosed publicly on CVE.org and NVD
2. 2026-10-26 11:10 UTC Last updatedMost recent change to the record

Details

Published

2026-06-18 23:16 UTC

Last modified

2026-10-26 11:10 UTC

Exploit probability

0.2% · EPSS, next 30 days

Weakness

CWE-863 · Incorrect authorization

Record status

Deferred

Attack profile CVSS 4.0 · 6.8

Attack vector

Local

Complexity

Low

Attack requirements

None

Privileges needed

None

User interaction

Passive

Confidentiality

High

Integrity

None

Availability

None