The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions to install the @qwencloud/qwencloud-cli package from the npm registry. This is a legitimate vendor tool associated with the skill's purpose and the identified author 'qwencloud'.\n- [COMMAND_EXECUTION]: The skill executes various commands through the qwencloud CLI, such as auth status, auth login, usage summary, and usage breakdown. These commands are necessary for the skill's core functionality and do not exhibit malicious behavior.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes structured data from the output of CLI commands to present summaries. While this represents a potential injection surface, the risk is inherent to the skill's purpose of interpreting tool output.\n
Ingestion points: Output from qwencloud usage commands (SKILL.md).\n
Boundary markers: The skill explicitly instructs the agent to separate CLI output from its analysis using the --- separator.\n
Capability inventory: Executes qwencloud CLI commands, installs npm packages, and opens browser URLs via system tools (open, xdg-open, start).\n
Sanitization: No explicit data sanitization is mentioned, but the skill follows a structured parsing approach using --format json.

qwencloud-usage