The Agent Skills Directory

[PROMPT_INJECTION]: No evidence of instructions attempting to bypass safety filters, override agent behavior, or extract system prompts. The skill uses natural instructional language for routing purposes.
[DATA_EXFILTRATION]: No sensitive file paths, hardcoded credentials, or network operations (e.g., curl, wget) were detected. The skill does not access or transmit data externally.
[REMOTE_CODE_EXECUTION]: No patterns of remote script execution, package installation, or dynamic code execution (e.g., eval, exec) are present.
[INDIRECT_PROMPT_INJECTION]: While the skill processes user input to determine task routing, it lacks the technical capabilities (file writing, network access, or command execution) to be exploited via data poisoning. There are no boundary markers or explicit sanitization, but the lack of executable capabilities renders the surface inert.
[COMMAND_EXECUTION]: The skill contains no shell commands, subprocess calls, or privilege escalation attempts. It functions entirely through high-level natural language instructions.

creative-production