CVE-2026-56835: PraisonAI authorization bypass in Slack app_mention handler

CVE-2026-56835 · Severity: high · CVSS 8.3 · Published 2026-10-23

PraisonAI is a multi-agent system that integrates with Slack. A flaw in its Slack bot allows unauthorized users to trigger the PraisonAI agent by mentioning the bot, even when those users or channels are blocked by the administrator's access controls. An attacker with access to any Slack channel where the bot is installed can invoke the agent and potentially extract sensitive information or execute unwanted actions.

Technologies: MervinPraison PraisonAI, praisonai (PyPI). Vendors: MervinPraison, PyPI.

Quick answers

Is CVE-2026-56835 exploited in the wild?
No. Junglewise Threat Intelligence has no report of CVE-2026-56835 being exploited in the wild.
How severe is CVE-2026-56835?
High severity, with a CVSS score of 8.3, published 23 October 2026.
Which products and versions are affected?
MervinPraison PraisonAI 3.11.0 to 4.6.58; praisonai (PyPI) ECOSYSTEM:introduced=3.11.0,fixed=4.6.59.
Is there a fix for CVE-2026-56835?
Yes. praisonai (PyPI): upgrade to 4.6.59 or later. Junglewise finds every place you run it, drafts the upgrade and applies it once you approve.
How is CVE-2026-56835 attacked?
It is attacked over the network.
Am I affected by CVE-2026-56835?
Check whether you run MervinPraison PraisonAI or praisonai (PyPI) at an affected version. Junglewise does that check for you around the clock: connect your code, cloud and domains, and it matches new vulnerabilities like this one against what you actually run and tells you when one reaches you.

How to fix

  • praisonai (PyPI): upgrade to 4.6.59 or later
    pip install "praisonai>=4.6.59"

Technical details

The Slack bot implements user and channel authorization checks in its message event handler but fails to apply the same checks in the app_mention event handler. When a user mentions the bot, the app_mention path directly invokes the agent session without calling is_channel_allowed(), is_user_allowed(), or UnknownUserHandler.handle(). This allows attackers to bypass allowed_users, allowed_channels, and unknown_user_policy restrictions that are properly enforced on regular messages.

Affected products

  • MervinPraison PraisonAI 3.11.0 to 4.6.58

Identifiers

  • CVE-2026-56835
  • GHSA-qvpf-j64c-jmhr
  • PYSEC-2026-3518

Timeline

  • 2026-06-17: disclosed
  • 2026-10-23: advisory

References

This page as Markdown. Free to reuse under CC BY 4.0 with a link back.