arkcli-deploy
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill manages the execution of the
arkclicommand-line interface for tasks such as creating online inference endpoints (+deploy) and checking authentication status. It includes specific technical requirements for these commands, such as using PascalCase for JSON-based flags and managing environment-specific behaviors (e.g., TTY vs. non-TTY). - [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it processes user-provided keywords, model IDs, and endpoint names to parameterize CLI commands. To mitigate this, the instructions mandate that the agent must verify all candidates against real-time, structured output from the tool and obtain explicit user confirmation for any billable write operations.
Audit Metadata