The Agent Skills Directory

[SAFE]: The skill is composed entirely of natural language instructions and metadata configuration. No executable code, scripts, or dangerous patterns were identified during the analysis.- [PROMPT_INJECTION]: The skill provides instructions to rephrase the last agent response into simpler language. While this modifies agent behavior, it does so within the intended functional scope and does not attempt to bypass safety guidelines, extract system prompts, or override core constraints.- [DATA_EXFILTRATION]: The skill does not contain any network operations, external URL references, or commands to access sensitive system files or credentials.- [REMOTE_CODE_EXECUTION]: No package manifests (e.g., package.json, requirements.txt) or remote download commands (e.g., curl, wget) are present. The skill does not perform any dynamic code execution.

bro