The Agent Skills Directory

[SAFE]: The skill uses multi-agent orchestration entirely within the local execution environment for text synthesis and analysis. It does not involve any dangerous code execution, remote script fetching, or unauthorized file system operations. The subagents described are internal LLM role-play instances aimed at diverse diagnosis generation.
[SAFE]: All referenced files are local markdown documents containing theoretical frameworks and illustrative examples. There are no external downloads from untrusted sources; the tool recommendations in the documentation are links to other skills within the same environment.
[SAFE]: The skill includes instructions for data retrieval (Stage 2) to ground claims in evidence, but this is scoped to the agent's available tools (like web search or internal file reading) and does not involve exfiltrating sensitive local configuration files or credentials.
[SAFE]: No obfuscation, prompt injection, or persistence mechanisms were detected. The skill follows best practices by providing a clear 'Bad-strategy audit' to ensure the AI's output is concrete and faces the actual strategic challenge provided by the user.

strategy-kernel