The Agent Skills Directory

[SAFE]: The skill documentation focuses on software design principles such as 'deep modules', 'seams', and 'leverage'. Analysis of all provided files shows no evidence of malicious instructions, credential harvesting, or unauthorized system access.
[NO_CODE]: The skill does not bundle any scripts or executable files; it functions as a set of instructions and conceptual frameworks for the agent to use during architectural discussions.
[INDIRECT_PROMPT_INJECTION]: While the skill is intended to process and refactor user-provided codebases (an untrusted ingestion point), it does not possess any capabilities like file writing, network access, or command execution that could be leveraged by an indirect prompt injection attack.

codebase-design