terminal-opener
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill's core function is to launch arbitrary executables and arguments provided by the user. It mitigates shell injection attacks by using
child_process.spawnwithshell: falseand implementing custom validation logic for the executable name and arguments. - [DATA_EXPOSURE]: The script inherits the calling process's environment variables without filtering. This behavior is documented in
SKILL.md, noting that secrets in the environment will be accessible to the launched command. - [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection if it is used to execute commands derived from untrusted external data (e.g., instructions found in a web page or file).
- Ingestion points: Arguments passed to
open-terminal.jsvia the agent. - Boundary markers: None present in the instructions.
- Capability inventory:
child_process.spawnSyncandchild_process.spawninscripts/open-terminal.jsused to launch external processes. - Sanitization: Extensive validation in
scripts/open-terminal.jsensures that the input is treated as a single executable and a list of arguments, preventing shell string interpolation and injection.
Audit Metadata