The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes rg (ripgrep) and grep to search through the local repository's models and migrations. This is a standard procedure for identifying existing code patterns and relationships within a Rails application.
[COMMAND_EXECUTION]: The skill is configured to invoke bin/rails for generating boilerplate, running migrations, and opening the Rails console. These actions are explicitly restricted in the YAML frontmatter using wildcards (e.g., bin/rails generate*), which limits the agent to a specific set of safe development commands.
[DATA_EXPOSURE]: The skill reads db/schema.rb and migration files. While these files describe the application's data structure, access is restricted to the local environment and is essential for the skill's primary purpose of managing database associations.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from the local filesystem, such as source code and schema files. This represents a potential surface for indirect prompt injection if an attacker can influence the codebase, but the risk is inherent to development tools and is mitigated by the skill's narrow focus on Active Record logic.

active-record-associations