---
title: "CVE-2026-47391: PraisonAI's unauthenticated A2A official example can reach real LLM-driven `eval()` tool execution"
url: https://erroneo.us/security/cve/CVE-2026-47391/
description: "CVE-2026-47391 (PraisonAI (PyPI)): PraisonAI is a multi-agent teams system. Prior to version 4.6.40, PraisonAI's first-party A2A server example exposes an…"
type: vulnerability-record
section: "Security Digest"
date_published: 2026-07-21
date_modified: 2026-07-23
publisher: erroneo.us
author: erroneo.us Security Desk
cve: "CVE-2026-47391"
kev: false
cvss: 9.8
epss: 0.01165
license: All rights reserved. Quote with attribution and a link.
---

# CVE-2026-47391: PraisonAI's unauthenticated A2A official example can reach real LLM-driven `eval()` tool execution

**Signal:** Critical and remotely reachable. No known exploitation yet, but it is network-reachable without credentials. (Computed from CISA KEV, EPSS and CVSS; not an editorial verdict.)

## Should you act?

- Not in CISA KEV , but CISA’s SSVC assessment records a public proof-of-concept, so attack code exists.
- EPSS 1.2% , percentile 67: FIRST’s model estimates a 1.2% chance of exploitation activity in the next 30 days, higher than 67% of all scored CVEs (moderate). EPSS is a probability, not a verdict.
- Remotely reachable with no login and no user interaction. If the vulnerable service is exposed to the internet, prioritize it.
- Only applies if your dependency tree includes PraisonAI (PyPI) <= 4.6.39, fixed in 4.6.40 . Check your lockfile, including transitive dependencies.

## What it is

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.

## Affected packages

- PyPI `PraisonAI` <= 4.6.39 → fixed: 4.6.40

## Affected products (CVE record)

- MervinPraison PraisonAI: < 4.6.40

## Severity

- GitHub_M CVSS 3.1: 9.8 `CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H`
- GitHub CVSS 3.1: 9.8 `CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H`

## Exploitation

- CISA KEV: not listed (no confirmed in-the-wild exploitation known to CISA).
- CISA SSVC assessment: exploitation poc , automatable yes , technical impact total .
- FIRST EPSS (model of 2026-10-21): 1.17% probability of exploitation activity in the next 30 days; percentile 67.

## Sources

- [GitHub Advisory GHSA-vg22-4gmj-prxw](https://github.com/advisories/GHSA-vg22-4gmj-prxw)
- [CVE.org record CVE-2026-47391](https://www.cve.org/CVERecord?id=CVE-2026-47391)
- [NVD entry CVE-2026-47391](https://nvd.nist.gov/vuln/detail/CVE-2026-47391)
- [github.com](https://github.com/MervinPraison/PraisonAI/security/advisories/GHSA-vg22-4gmj-prxw)
- [github.com](https://github.com/MervinPraison/PraisonAI/pull/1793)
- [github.com](https://github.com/MervinPraison/PraisonAI/commit/e0fb8e7dd1ee6759c18ed07f436c21dbd9c20747)

---
Source page: https://erroneo.us/security/cve/CVE-2026-47391/ · Security Digest: https://erroneo.us/security/ · More: https://erroneo.us/llms.txt