script-execute
Warn
Audited by Gen Agent Trust Hub on Aug 21, 2026
Risk Level: MEDIUMREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill utilizes the Roslyn compiler to execute C# code provided in the csharpCode parameter, enabling arbitrary logic to be run at runtime.
- [REMOTE_CODE_EXECUTION]: The capability to execute arbitrary code provides a potential pathway for unauthorized file system access, network operations, or system command execution, depending on the host process permissions.
- [INDIRECT_PROMPT_INJECTION]: The tool exposes a significant attack surface for indirect prompt injection. An agent could be manipulated into executing malicious code if it processes instructions or scripts from untrusted external data sources.
- Ingestion points: The csharpCode input parameter defined in SKILL.md.
- Boundary markers: Absent; the skill does not implement mechanisms to distinguish between safe developer-provided code and potentially malicious injected code.
- Capability inventory: Full C# execution capabilities, which may include access to system-level APIs and namespaces.
- Sanitization: Absent; the provided code is compiled and executed without security validation or sandboxing.
Audit Metadata