hubspot-create-meeting
Warn
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION] (MEDIUM): The skill invokes Python scripts residing in a system directory (
00-system/skills/hubspot/hubspot-master/scripts/). This allows the agent to execute code on the host machine via the command line. - [PROMPT_INJECTION] (MEDIUM): User-provided strings for the
--titleand--bodyarguments are interpolated into shell commands. This pattern is vulnerable to shell command injection (e.g., using semicolons or backticks) if the agent does not strictly sanitize the input before execution. - [INDIRECT PROMPT INJECTION] (MEDIUM): 1. Ingestion points: User inputs for the title and body of the meeting. 2. Boundary markers: None. No escaping or delimiters are defined in the command template. 3. Capability inventory: Execution of system commands via
pythonscripts. 4. Sanitization: None visible in the skill manifest.
Audit Metadata