The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to analyze, review, and refactor Rails application code provided by the user, which presents an attack surface for instructions embedded within that data to influence the agent's behavior.
Ingestion points: Project source files identified by path patterns such as app/models/**/*.rb, app/controllers/**/*_controller.rb, spec/**/*_spec.rb, and linter configuration files like .rubocop.yml (SKILL.md).
Boundary markers: The instructions lack explicit delimitation or directives for the agent to ignore potentially malicious natural language instructions hidden in code comments or string literals within the files being processed.
Capability inventory: The skill directs the agent to execute shell commands to run linters and to generate new code implementation based on the content of the analyzed files (SKILL.md).
Sanitization: There is no evidence of input validation, filtering, or sanitization of the processed source code before it is incorporated into the agent's context for refactoring.
[COMMAND_EXECUTION]: The skill workflow requires the agent to detect and execute local command-line tools for linting and code verification.
Evidence: Step 1 in the "Core Process" section of SKILL.md explicitly instructs the agent to "Run linter" using detected tools such as rubocop or standard.

apply-code-conventions