The Agent Skills Directory

[SAFE]: Analysis of the skill instructions and processes revealed no security violations. The use of cargo xtask and git is appropriate for the tool's function as a project management assistant within the developer's environment.
[COMMAND_EXECUTION]: The skill executes local commands such as cargo xtask change new and git push. These are standard development tasks used to scaffold documentation files and interact with the project's version control repository.
[PROMPT_INJECTION]: The skill processes user input for feature ideas and branch names. While this provides an ingestion point for external data, the skill is configured with disable-model-invocation: true, which means the procedural steps are executed directly without passing the user input to an LLM for logic determination, significantly mitigating risks of indirect prompt injection.

propose