The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill downloads and installs its core logic from a third-party GitHub repository (https://github.com/tianyilt/qzcli_tool).\n- [COMMAND_EXECUTION]: The skill requires the use of the Bash(*) tool to perform environment setup, job submission, and interactions with the Qizhi platform.\n- [CREDENTIALS_UNSAFE]: The skill manages authentication credentials, including usernames and passwords, by storing them in local environment files (~/.qzcli/.env) and configuration files (~/.qzcli/config.json).\n- [PROMPT_INJECTION]: The skill processes job statuses, resource names, and workspace details from the external Qizhi platform, creating a surface for potential indirect prompt injection attacks.\n
Ingestion points: Data from the shared platform is ingested via commands such as qzcli ls, qzcli status, and qzcli res.\n
Boundary markers: No specific delimiters or instructions to ignore embedded commands are included in the prompt templates.\n
Capability inventory: The agent has extensive capabilities, including full shell access (Bash(*)) and file operations (Read, Write).\n
Sanitization: There is no evidence of validation or sanitization of data retrieved from the remote platform before it is presented to the agent.

qzcli