byted-viking-cli

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads an installation script for the 'viking-cli' tool from 'viking-skills.tos-cn-beijing.volces.com', which is a domain associated with the vendor's cloud platform (Volcengine).
  • [PRIVILEGE_ESCALATION]: The documentation provides an installation command using 'sudo' to allow the installation of the CLI tool into restricted system directories like '/usr/local'.
  • [COMMAND_EXECUTION]: The skill operates by translating natural language user requests into executable shell commands using the 'viking-cli' binary.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input to populate command arguments such as query strings and metadata.
  • Ingestion points: User queries provided to the agent are interpolated into CLI flags like --query, --content, and --desc.
  • Boundary markers: No specific delimiters or safety instructions are used to separate user data from the command structure.
  • Capability inventory: The skill possesses the ability to execute shell commands, perform network API calls via the CLI, and read/write data to the vendor's cloud databases.
  • Sanitization: There is no evidence of string sanitization or validation of user-provided content before it is passed to the command line.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 03:15 PM