CVE-2026-57137: Auth Bypass (CVSS 8.8)

Description

PraisonAI is a multi-agent teams system. From 1.4.0 until 1.7.2, createAgentLoop() in src/praisonai-ts/src/ai/agent-loop.ts passes executable tools to generateText() before invoking the onToolCall approval callback. Because the wrapped AI SDK executes tool handlers during generation, a callback that returns false records tool_rejected only after the denied tool has already produced side effects and populated toolResults. Applications using onToolCall as a human or policy approval boundary can therefore execute rejected file, command, API, or data-modifying operations. This issue is fixed in version 1.7.2.

Is your site exposed to CVE-2026-57137?

Run a free security scan — no signup, results in seconds.

CVSS v3.1 Score

8.8

HIGH

CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H

EPSS — Exploit Prediction

0.0036

Probability of exploitation

0.30%

Percentile rank

EPSS estimates the probability that this vulnerability will be exploited in the wild within the next 30 days. A higher score means more likely to be exploited.

Weakness Type (CWE)

CWE-693 CWE-693

CWE-862 Missing Authorization

CWE-863 Incorrect Authorization

References

Other References

https://github.com/MervinPraison/PraisonAI/commit/6b4b59fbe71a74983626c7a0cf7e10c47cd004f9 https://github.com/MervinPraison/PraisonAI/releases/tag/v4.6.62 https://github.com/MervinPraison/PraisonAI/security/advisories/GHSA-h2w2-v7j6-xqm4 https://github.com/MervinPraison/PraisonAI/security/advisories/GHSA-h2w2-v7j6-xqm4

Frequently Asked Questions

What is CVE-2026-57137? +

PraisonAI is a multi-agent teams system. From 1.4.0 until 1.7.2, createAgentLoop() in src/praisonai-ts/src/ai/agent-loop.ts passes executable tools to generateText() before invoking the onToolCall approval callback. Because the wrapped AI SDK executes tool handlers during generation, a callback that returns false records tool_rejected only after the denied tool has already produced side effects and populated toolResults. Applications using onToolCall as a human or policy approval boundary can therefore execute rejected file, command, API, or data-modifying operations. This issue is fixed in version 1.7.2. It has a CVSS v3.1 base score of 8.8 (HIGH).

How severe is CVE-2026-57137? +

CVE-2026-57137 has a CVSS v3.1 score of 8.8 out of 10, rated HIGH. This is a high-severity vulnerability that should be prioritized for patching. The EPSS score is 0.0036, placing it in the 0th percentile for exploitation probability.

How do I check if I'm vulnerable to CVE-2026-57137? +

You can use Secably's free Website Scanner to check your website for known vulnerabilities. For infrastructure scanning, use the Port Scanner to identify exposed services that may be affected. Check the vendor advisories linked above for specific patch and version information.

Related Vulnerabilities

CVE-2026-48033

Hulumi is an open-source toolkit that ships secure-by-default cloud and platform infrastructure components for Pulumi. Prior to version 1.4.0, policy …

CVE-2026-0278

Multiple protection mechanism failures in the Prisma Access Agent Data Loss Prevention (DLP) component for Windows allow a local user …

CVE-2025-3114

Code Execution via Malicious Files: Attackers can create specially crafted files with embedded code that may execute without adequate security …

CVE-2024-51481

Nix is a package manager for Linux and other Unix systems. On macOS, built-in builders (such as `builtin:fetchurl`, exposed to …

CVE-2025-65100

Isar is an integration system for automated root filesystem generation. In versions 0.11-rc1 and 0.11, defining ISAR_APT_SNAPSHOT_DATE alone does not …

CVE-2025-29864

Protection Mechanism Failure vulnerability in ESTsoft ALZip on Windows allows SmartScreen bypass.This issue affects ALZip: from 12.01 before 12.29.

CVE-2026-57137