The Agent Skills Directory

[Prompt Injection] (SAFE): The skill uses natural instructional language to define a persona. There are no attempts to bypass safety filters, extract system prompts, or override core agent instructions.
[Data Exposure & Exfiltration] (SAFE): No network operations (curl, wget, fetch) or sensitive file path accesses were detected. The skill is entirely local and conversational.
[Unverifiable Dependencies & Remote Code Execution] (SAFE): There are no package managers (npm, pip) used, no remote scripts downloaded, and no dynamic code execution patterns (eval, exec) present.
[Indirect Prompt Injection] (LOW): The skill is designed to ingest and process user-provided research context. While this is an ingestion surface for potential indirect instructions, the skill lacks any 'capabilities' (such as file-writing, network access, or command execution) that could be exploited by such an injection.
[Persistence & Privilege Escalation] (SAFE): No commands for modifying system startup, shell profiles, or acquiring elevated permissions were found.

scientific-brainstorming