CVE-2026-47391: Code Injection (CVSS 9.8)

Description

PraisonAI is a multi-agent teams system. Prior to version 4.6.40, PraisonAI's first-party A2A server example exposes an unauthenticated A2A JSON-RPC endpoint and registers a `calculate(expression)` tool implemented with Python `eval()`. The example also binds to `0.0.0.0`. A remote unauthenticated attacker can send `message/send` to `/a2a`; the request reaches `agent.chat()`, and a real LLM can invoke the registered `calculate` tool. In testing with `gemini/gemini-2.5-flash-lite`, this resulted in arbitrary Python execution in the server process, confirmed by creation of a marker file from an unauthenticated HTTP request. The issue affects deployments following the official A2A example or similar unauthenticated public A2A deployments with unsafe tools. The default unauthenticated A2A surface also exposes task history and task cancellation APIs, increasing confidentiality and integrity impact. Version 4.6.40 patches the issue.

Is your site exposed to CVE-2026-47391?

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

CVSS v3.1 Score

9.8

CRITICAL

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

Weakness Type (CWE)

CWE-95 CWE-95

CWE-306 Missing Authentication

References

Other References

https://github.com/MervinPraison/PraisonAI/commit/e0fb8e7dd1ee6759c18ed07f436c21dbd9c20747 https://github.com/MervinPraison/PraisonAI/pull/1793 https://github.com/MervinPraison/PraisonAI/security/advisories/GHSA-vg22-4gmj-prxw

Frequently Asked Questions

What is CVE-2026-47391? +

PraisonAI is a multi-agent teams system. Prior to version 4.6.40, PraisonAI's first-party A2A server example exposes an unauthenticated A2A JSON-RPC endpoint and registers a `calculate(expression)` tool implemented with Python `eval()`. The example also binds to `0.0.0.0`. A remote unauthenticated attacker can send `message/send` to `/a2a`; the request reaches `agent.chat()`, and a real LLM can invoke the registered `calculate` tool. In testing with `gemini/gemini-2.5-flash-lite`, this resulted in arbitrary Python execution in the server process, confirmed by creation of a marker file from an unauthenticated HTTP request. The issue affects deployments following the official A2A example or similar unauthenticated public A2A deployments with unsafe tools. The default unauthenticated A2A surface also exposes task history and task cancellation APIs, increasing confidentiality and integrity impact. Version 4.6.40 patches the issue. It has a CVSS v3.1 base score of 9.8 (CRITICAL).

How severe is CVE-2026-47391? +

CVE-2026-47391 has a CVSS v3.1 score of 9.8 out of 10, rated CRITICAL. This is a critical vulnerability that should be patched immediately.

How do I check if I'm vulnerable to CVE-2026-47391? +

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-2026-8914

In Teltonika Networks RUTOS devices, running versions 7.22 through 7.23.2 and TSWOS devices running versions 1.09 through 1.09.1, due to …

CVE-2026-44128

SEPPmail Secure Email Gateway before version 15.0.2.1 allows unauthenticated remote code execution in the new GINA UI because an endpoint …

CVE-2025-0868

A vulnerability, that could result in Remote Code Execution (RCE), has been found in DocsGPT. Due to improper parsing of …

CVE-2025-4318

The AWS Amplify Studio UI component property expressions in the aws-amplify/amplify-codegen-ui package lack input validation. This could potentially allow an …

CVE-2025-47271

The OZI action is a GitHub Action that publishes releases to PyPI and mirror releases, signature bundles, and provenance in …

CVE-2025-49598

conda-forge-ci-setup is a package installed by conda-forge each time a build is run on CI. The conda-forge-ci-setup-feedstock setup script is …

CVE-2026-47391