The Agent Skills Directory

[SAFE]: The skill employs standard development tools including git, ls, find, and wc to perform architectural analysis. These commands are used for their intended purpose of inspecting file structures and metrics.
[SAFE]: External dependencies and references mentioned (e.g., numpy, scipy, sklearn, sqlite-vec, and the cashew library) are well-known technology components and official open-source projects.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted source code from external repositories.
Ingestion points: Code is ingested via git clone from user-provided URLs in SKILL.md.
Boundary markers: The skill explicitly mandates a "clean-room" approach, instructing the agent to never include source code snippets from the reference codebase in the output.
Capability inventory: The skill uses basic file system read operations; it does not install unknown packages or execute the code being analyzed.
Sanitization: Instructions require the agent to abstract the system into natural language design principles and Mermaid diagrams, which prevents the direct propagation of potentially malicious instructions embedded in comments or code.

software-architecture-analysis