eightctl
Warn
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill metadata contains installation instructions for an external binary using 'go install github.com/steipete/eightctl/cmd/eightctl@latest'. The source repository is owned by a third-party user and is not part of a trusted organization list.
- [COMMAND_EXECUTION]: The skill's primary functionality relies on executing the 'eightctl' command-line interface on the local system to interact with hardware.
- [CREDENTIALS_UNSAFE]: The skill requires the configuration of sensitive authentication data, including 'EIGHTCTL_EMAIL' and 'EIGHTCTL_PASSWORD' environment variables and a local configuration file at '~/.config/eightctl/config.yaml'. Accessing these credentials is a necessary step for the tool to function but exposes sensitive user data to the environment.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from tool outputs without safety boundaries. * Ingestion points: Tool output from commands such as 'eightctl status' and 'eightctl alarm list'. * Boundary markers: Absent. * Capability inventory: Local subprocess execution of the 'eightctl' binary. * Sanitization: No sanitization or validation of the tool's output is performed before it is returned to the agent context.
Audit Metadata