The Agent Skills Directory

[SAFE]: The skill's primary purpose is context orchestration and stack auditing. It explicitly enforces high-integrity practices by requiring verification of technology ownership against official vendor documentation rather than relying on search metrics or popularity.
[SAFE]: The skill implements strict boundary controls by refusing to install community-made context without explicit user approval and recording gaps when official context is unavailable, preventing the silent substitution of untrusted third-party instructions.
[SAFE]: Interaction with external sources like llms.txt is handled as dynamic retrieval rather than static file copying, ensuring that instruction surfaces like AGENTS.md remain linked to live vendor updates rather than becoming stale or hosting uncontrolled code.
[SAFE]: The tool respects user authorship and environment integrity by preserving existing symbolic links and user-defined rules during the context injection process.

add-stack-context