The Agent Skills Directory

[SAFE]: The skill is primarily a collection of design patterns and educational references for building complex systems. The instructions and documentation do not contain any malicious patterns or attempts to bypass safety filters.
[EXTERNAL_DOWNLOADS]: The Python code snippets in references/metacognitive-patterns.md reference the numpy library for statistical calculations (calibration error). This is a well-known, legitimate library for scientific computing and is documented here as part of a logic example.
[COMMAND_EXECUTION]: The skill contains various Python and Haskell code snippets across its reference files. These are provided as conceptual examples (e.g., self-modifying dictionaries, Y combinators, and search algorithms) rather than instructions for the agent to execute arbitrary system commands or download external payloads.

recursive-systems-architect