CVE-2026-56838

The record

PraisonAI is a multi-agent teams system. From version 4.5.87 to before version 4.6.59, PraisonAI recipe execution blocks default-denied dangerous tools unless the caller explicitly passes allow_dangerous_tools=True. The normal recipe.run() path enforces this with _check_tool_policy(). The streaming path, recipe.run_stream(), loads the same recipe, checks dependencies, and then calls _execute_recipe() without running the dangerous-tool policy check. As a result, a recipe that honestly declares execute_command in TEMPLATE.yaml requires.tools is denied by recipe.run(), but reaches the execution engine through recipe.run_stream() with allow_dangerous_tools=False. This issue has been patched in version 4.6.59.

Affected packages

Related vulnerabilities

• CVE-2026-62853 ThingLinks is an enterprise IoT platform. CWE-78
• CVE-2026-119335 Pi-hole is a DNS sinkhole that protects devices from unwanted content without installing… CWE-78
• CVE-2026-116241 pgAdmin 4's Import/Export Data tool wraps a user's query export in a psql \copy (...) TO… CWE-78
• CVE-2026-119713 Paperclip 2026.403.0 before 2026.416.0 contains a remote code execution vulnerability… CWE-78
• CVE-2026-119288 Zuke is a code-first, strongly typed build automation system for Deno. CWE-78
• CVE-2026-118238 The Tenda W18E web management server is vulnerable to OS command injection via the… CWE-78
• CVE-2026-91913 Dell OpenManage Enterprise, versions prior to 4.8.0, contains an Improper Neutralization… CWE-78
• CVE-2026-118992 CallBackery versions before 0.62.0 for Perl allow OS command execution via sqlite3 shell… CWE-78