CVE-2026-30306
Prompt Injection Vulnerability in SakaDev Enables Arbitrary Command Execution
Publication date: 2026-03-30
Last updated on: 2026-04-08
Assigner: MITRE
Description
Description
CVSS Scores
EPSS Scores
| Probability: | |
| Percentile: |
Meta Information
Affected Vendors & Products
| Vendor | Product | Version / Range |
|---|---|---|
| rahmanazhar | sakadev | to 4.0.6 (exc) |
Helpful Resources
Exploitability
| CWE ID | Description |
|---|---|
| CWE-94 | The product constructs all or part of a code segment using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the syntax or behavior of the intended code segment. |
Attack-Flow Graph
AI Powered Q&A
How does this vulnerability affect compliance with common standards and regulations (like GDPR, HIPAA)?:
The provided information does not specify how the vulnerability in SakaDev impacts compliance with common standards and regulations such as GDPR or HIPAA.
What immediate steps should I take to mitigate this vulnerability?
The vulnerability arises from the automatic terminal command execution feature in SakaDev, which misclassifies malicious commands as safe and executes them without user approval.
Immediate mitigation steps include disabling or avoiding the use of the 'Execute safe commands' mode that automatically runs commands deemed safe by the model, as this mode is susceptible to prompt injection attacks.
Additionally, updating SakaDev to a version beyond 4.0.5 (if available) or applying any patches provided by the maintainers can help mitigate the issue.
Until a fix is applied, always require manual approval for command execution to prevent arbitrary command execution.
Can you explain this vulnerability to me?
This vulnerability exists in SakaDev's design for automatic terminal command execution, which offers two modes: executing safe commands automatically and executing all commands with user approval for potentially destructive ones.
The system relies on a model to classify commands as safe or potentially destructive. However, an attacker can exploit this by using a generic template to wrap malicious commands, tricking the model into classifying them as safe.
As a result, the malicious commands are executed automatically without requiring user approval, leading to arbitrary command execution.
How can this vulnerability impact me? :
This vulnerability can lead to arbitrary command execution on systems using SakaDev's automatic terminal command execution feature.
An attacker can bypass user approval mechanisms and execute malicious commands, potentially compromising system integrity, confidentiality, and availability.
How can this vulnerability be detected on my network or system? Can you suggest some commands?
This vulnerability involves the automatic execution of terminal commands by the SakaDev Visual Studio Code extension, where malicious commands can be misclassified as safe and executed without user approval.
To detect this vulnerability on your system, you should monitor the execution of terminal commands initiated by the SakaDev extension, especially those that bypass user approval unexpectedly.
Since the vulnerability is related to prompt injection leading to arbitrary command execution, you can look for unusual or unexpected commands running in the terminal associated with VSCode processes.
Specific commands to detect suspicious activity might include:
- On Linux/macOS, use: `ps aux | grep code` to identify VSCode processes and monitor their child processes for unexpected commands.
- Use terminal command history inspection: `cat ~/.bash_history` or equivalent to look for suspicious commands executed by the user or by VSCode.
- Monitor real-time terminal activity with: `sudo auditctl -w /usr/bin/code -p x -k vscode_exec` (Linux audit framework) to log executions related to VSCode.
- Check VSCode extension logs or output panels for any automated command execution logs that might indicate bypassing of user approval.
Note that no specific detection commands are provided in the available resources, so these suggestions are general approaches based on the nature of the vulnerability.