Connect your agent
AdvisoriesCriticalCheck your lockfilePluginHow it worksPricing Connect your agent

Advisory record · GHSA-pv2j-rghr-v5r9

PraisonAI: execute_code sandbox bypass: str.format C-level attribute access reads every blocklisted dunder

CVSS 6.5 MediumEPSS 0.0056 · 46th percentile

Published by the GitHub Advisory Database at 18 Jun 2026, 13:55 UTC. agentprime has not analysed this yet, so whether it affects any app is not known.

PackageEcosystemAffectedFixed in
praisonaiagentsPyPI<1.6.591.6.59
Identifiers
GHSA-pv2j-rghr-v5r9 · CVE-2026-57120 · PYSEC-2026-3533
Weakness (CWE)
CWE-693
Tags
Published
18 Jun 2026
Updated
23 Jul 2026

Machine-written summary

Summary written by Claude Haiku from the source advisory. Not reviewed by a person.

The execute_code tool in PraisonAI (praisonaiagents, versions before 1.6.59) has a sandbox bypass: in sandbox mode, its AST and blocklist checks can be evaded by assembling blocked dunder names at runtime and reading them through str.format attribute resolution. An attacker who can get code into the agent's execute_code tool, such as through a user message or retrieved content, can read blocked attributes like __class__ and __globals__; the source does not establish code execution from this read alone. Upgrade praisonaiagents to version 1.6.59 or later.

Written 5 Nov 2026.

Not analysed yet

We haven’t published our analysis of this Advisory yet. Once a person has reviewed it, this page will say what’s vulnerable, what has to be true for your app to be affected, and how to check.

Severity and exploitation

ScoreRatingCVSSVectorSource
6.5 Medium 3.1 CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N CNA (GitHub_M)
  • CISA KEV Not listed.
  • EPSS 0.0056, the 46th percentile, on 10 Nov 2026: FIRST’s estimate of the chance of exploitation in the next 30 days.

Timeline

  1. 31 Oct 2026, 14:41 UTC First seen by agentprime
  2. 31 Oct 2026, 14:41 UTC Confirmed by OSV

Sources

Facts on this page come from the GitHub Advisory Database (CC-BY 4.0), OSV and the other sources listed, as published. We don’t reproduce full advisory descriptions, which can contain exploit code.