leverage-cloud-agents
Fail
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill downloads an installation script from the vendor's domain at https://app.leverage.computer/install/cli to a temporary file and executes it using the shell. This is a standard procedure for installing the tool's CLI.
- [EXTERNAL_DOWNLOADS]: Fetches the CLI installer and onboarding playbooks from the official vendor infrastructure.
- [COMMAND_EXECUTION]: The skill executes several shell commands to manage environment paths, install browser-based plugins, and interact with the leverage CLI to manage cloud sessions.
- [INDIRECT_PROMPT_INJECTION]: The skill reads and summarizes local session history to generate an onboarding plan, creating a surface where malicious content in previous logs could influence agent behavior. Ingestion points: Output of the leverage onboard --scan command in SKILL.md. Boundary markers: Not used in the prompt interpolation. Capability inventory: The skill can execute shell commands and manage cloud-based agent sessions. Sanitization: No explicit sanitization of the scanned session data is described.
Recommendations
- HIGH: Downloads and executes remote code from: https://app.leverage.computer/install/cli - DO NOT USE without thorough review
Audit Metadata