The Agent Skills Directory

[SAFE]: The skill follows research best practices and does not exhibit malicious behavior. It includes explicit instructions for the AI to protect user privacy and avoid generating fabricated information such as invented literature or results.
[SAFE]: The skill uses a modular architecture referencing local files for specialized research tasks (e.g., experiment design, venue adaptation). All referenced scripts and files appear to be internal tools for registry maintenance rather than malicious remote code execution triggers.
[SAFE]: The multi-skill handoff logic is designed to respect user constraints, specifically allowing users to disable or skip certain functionalities to maintain control over the research process.

ccf-idea-optimizer