Vulnerability · Published Jul 21, 2026 · Assigned by GitHub_M
CVE-2026-47391PraisonAI's unauthenticated A2A official example can reach real LLM-driven `eval()` tool execution
Critical 9.8
In plain English
A critical-severity missing authentication in PraisonAI from MervinPraison. Proof-of-concept exploit code is public. The chance it is exploited in the next 30 days is estimated at 1.5% (higher than 74% of all known vulnerabilities). Fixed versions are available - upgrade the affected package.
Official 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.
Affected packages & fixed versions
| Ecosystem | Package | Fixed in | Advisory |
|---|---|---|---|
| PyPI | praisonai | 4.6.40 | GHSA-vg22-4gmj-prxw |
| PyPI | praisonai | 4.6.40 | PYSEC-2026-478 |
Affected products (as reported)
| Vendor | Product | Versions |
|---|---|---|
| MervinPraison | PraisonAI | < 4.6.40 |
Weakness: CWE-95, CWE-306 · Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H