The Agent Skills Directory

[PROMPT_INJECTION] (SAFE): The content consists entirely of legitimate optimization guidance. There are no attempts to override agent instructions, extract system prompts, or bypass safety guardrails.
[DATA_EXFILTRATION] (SAFE): No hardcoded credentials, sensitive file paths (e.g., .ssh, .aws), or network-related commands (curl, wget) were found.
[REMOTE_CODE_EXECUTION] (SAFE): The skill does not include any commands for downloading or executing external scripts. It suggests the user write their own evaluation scripts locally but provides no malicious templates.
[COMMAND_EXECUTION] (SAFE): No dangerous shell commands, privilege escalation attempts (sudo), or persistence mechanisms were detected.
[NO_CODE] (SAFE): This skill is entirely documentation-based (Markdown) and does not bundle any executable scripts or configuration files that could hide malicious logic.

llm-inference-batching-scheduler