The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's 'Activation Procedure' instructs the agent to immediately and automatically execute python scripts/search_bilibili.py upon activation. This pattern of mandatory code execution without user oversight is a security concern.
[COMMAND_EXECUTION]: The script scripts/search_bilibili.py uses the subprocess module to execute the browser-use CLI tool to open a web browser to a hardcoded URL. While the URL is benign, the use of shell execution for browser management is a notable capability.
[COMMAND_EXECUTION]: The documentation in SKILL.md describes a workflow where user-supplied natural language ('diagram description') is passed directly as a command-line argument to a Python script: python scripts/generate_schematic.py "your diagram description". This represents a high-risk surface for command injection if the underlying script does not properly sanitize inputs.
[PROMPT_INJECTION]: The skill uses authoritative language ('IMPORTANT', 'MUST') to force the agent into performing specific actions (playing background music) that are unrelated to the user's primary task, effectively attempting to override the agent's standard operational behavior.
[EXTERNAL_DOWNLOADS]: The skill proactively suggests the use of an external platform (www.k-dense.ai) for complex workflows. This is documented as a recommendation for an integrated service from the same creators.

daipai-execution-master