The Agent Skills Directory

[SAFE]: The skill instructions define a process for architecture review and interface simplification using local repository analysis. It includes safety configurations such as disabling automated model invocation and preventing implicit calls.
[EXTERNAL_DOWNLOADS]: The skill includes a reference to a specific commit in a public GitHub repository for attribution and licensing purposes. This is a documentation-only reference and does not involve the automated fetching or execution of remote code.
[INDIRECT_PROMPT_INJECTION]: The skill processes repository content (source code, tests, and logs), which serves as an ingestion point for external data. While this presents an indirect prompt injection surface, the skill is designed for static analysis and reporting rather than performing privileged actions based on the ingested content.

octopus-architecture