CVE-2026-57136
PraisonAI is a multi-agent teams system. From 1.2.3 until 1.7.2, CommandValidator in src/praisonai-ts/src/cli/features/sandbox-executor.ts v
CVSS
8.8
High
EPSS
—
KEV
—
Exploit Today
0
0-100
Published: Sep 15, 2026 · Last modified: Sep 15, 2026 · CWE-78 · CWE-693 · CWE-863
Not enough EPSS history yet.
PraisonAI is a multi-agent teams system. From 1.2.3 until 1.7.2, CommandValidator in src/praisonai-ts/src/cli/features/sandbox-executor.ts validates only the first whitespace-delimited executable against allowedCommands, then SandboxExecutor passes the complete command string to sh -c. A command beginning with an allowed executable can append a non-allowlisted command through shell metacharacters, causing arbitrary commands to run with the PraisonAI process privileges. This issue is fixed in version 1.7.2.
CVECVSSEPSSKEVRExploitTitleMod.
CVE-2026-919989.9 CRI—
——0Casdoor through 4.4.0 contains an authorization bypass vulnerability in the /api/mcp endpoint that allows attackers with any application's clientId and clientSecret to gain unrestricted access to user administration across all organizations. Attackers can enumerate user records including password salts and email addresses, create administrator accounts, modify existing users, and delete them in any organization by supplying legitimate credentials from a single application.1hCVE-2026-89308——
——0An unauthenticated OS command injection vulnerability exists in the ping.php endpoint, allowing remote attackers to execute arbitrary commands on the underlying operating system and achieve remote code execution.1hCVE-2026-571389.9 CRI—
——0PraisonAI is a multi-agent teams system. From 1.4.0 until 1.7.2, codeMode in src/praisonai-ts/src/tools/builtins/code-mode.ts executes untrusted JavaScript with new Function() inside with(sandbox) and relies on a small source-code blocklist plus shadowed process and require properties. Code can use ({}).constructor.constructor to recover the real Function constructor, obtain process and process.mainModule.require, and reach host filesystem and subprocess APIs despite the advertised sandbox. Attackers who control codeMode input can read secrets, modify files, execute commands, or exhaust the host process. This issue is fixed in version 1.7.2.2hCVE-2026-571378.8 HIG—
——0PraisonAI is a multi-agent teams system. From 1.4.0 until 1.7.2, createAgentLoop() in src/praisonai-ts/src/ai/agent-loop.ts passes executable tools to generateText() before invoking the onToolCall approval callback. Because the wrapped AI SDK executes tool handlers during generation, a callback that returns false records tool_rejected only after the denied tool has already produced side effects and populated toolResults. Applications using onToolCall as a human or policy approval boundary can therefore execute rejected file, command, API, or data-modifying operations. This issue is fixed in version 1.7.2.2hCVE-2026-571357.6 HIG—
——0PraisonAI is a multi-agent teams system. From 1.2.3 until 1.7.2, SandboxExecutor network-isolated mode in src/praisonai-ts/src/cli/features/sandbox-executor.ts uses buildEnv() only to inject invalid http_proxy and https_proxy environment variables and does not establish an operating-system network boundary. Programs that ignore those proxy variables can open sockets directly, allowing supposedly isolated commands to reach localhost, internal services, cloud metadata, or external hosts and potentially exfiltrate data. An initial remediation was released in version 1.7.2.2hCVE-2026-571348.2 HIG—
——0PraisonAI is a multi-agent teams system. From 1.5.1 until 1.7.2, MCPSecurity.evaluatePolicy() in src/praisonai-ts/src/mcp/security.ts invokes the configured credential validator only when AuthMethod is api-key or bearer. Basic and OAuth policies accept any non-empty Authorization header without calling auth.validate(), then return an authenticated result, allowing callers with invalid credentials to access MCP tools and resources protected by those policies. This issue is fixed in version 1.7.2.2h