The Agent Skills Directory

[SAFE]: The skill provides structured problem-solving methodologies (Five Whys, TRIZ, Systems Thinking, etc.) through text-based documentation. No malicious instructions or bypass attempts were found.
[NO_CODE]: The entire skill package consists of markdown files for reference. There are no Python scripts, Node.js files, or Shell scripts that could perform automated actions on the host system.
[REMOTE_CODE_EXECUTION]: No remote execution patterns are present. The skill does not utilize tools like curl or wget for runtime execution. The installation instructions in the README are for manual setup and point to the author's own repository.
[DATA_EXFILTRATION]: There is no evidence of commands or logic that access sensitive local environment files (e.g., .ssh, .aws, .env) or exfiltrate data to external servers.
[PROMPT_INJECTION]: The instructions are designed to guide the agent in logical analysis and do not contain directives to ignore safety filters or override system constraints.

problem-solver