The Agent Skills Directory

[PROMPT_INJECTION]: The skill uses a roleplay framework ('Doctor Strange') to facilitate simulation logic. It does not contain instructions aimed at bypassing safety guidelines or overriding core agent constraints.- [EXTERNAL_DOWNLOADS]: The skill utilizes 'WebSearch' and 'WebFetch' tools to populate an intelligence briefing for its simulations. These actions are limited to information retrieval and do not involve downloading executable scripts or packages from untrusted sources.- [DATA_EXFILTRATION]: Simulation outcomes are recorded in local memory files within the project directory. The skill does not access sensitive system paths (such as .ssh or .aws) or demonstrate patterns of exfiltrating data to external network endpoints.- [COMMAND_EXECUTION]: No shell commands, privilege escalation attempts, or persistent system modifications were detected. The skill operates within the boundaries of standard agentic tool interactions.

doctor-strange