cli-commands
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides an extensive reference for the
wmillcommand-line interface. These commands allow for full management of Windmill environments, including the execution of scripts and flows (script run,flow run), synchronization of local and remote workspaces (sync push,sync pull), and management of worker groups. - [INDIRECT_PROMPT_INJECTION]: The skill documents several commands that ingest and display data from external or remote sources, creating a potential attack surface for indirect prompt injection if the agent processes malicious instructions embedded in that data.
- Ingestion points: The agent receives external data through command outputs such as
job logs,object-storage preview,audit get, andscript show. - Boundary markers: There are no instructions defining specific delimiters or isolation techniques to distinguish CLI output from agent instructions.
- Capability inventory: The CLI provides powerful capabilities including code execution (
script run), credential management (token create), and workspace modification (sync push), which could be targeted by an injection attack. - Sanitization: The documentation does not specify methods for sanitizing or validating the data returned from these CLI commands before the agent processes it.
Audit Metadata