The Agent Skills Directory

[SAFE]: The skill serves as a configuration and instructional guide for organizing multi-agent tasks. It does not include any scripts, executable binaries, or system-level commands.
[SAFE]: No patterns of prompt injection, data exfiltration, or obfuscation were identified. The 'magic words' and activation phrases are intended as natural language triggers for agent orchestration rather than malicious overrides.
[SAFE]: The skill mentions integration with standard platform components like MCP servers and the Agent SDK but does not perform any unauthorized access or configuration changes.

cacos