The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute local development tools via npm run typecheck, npm run lint:all, and npm test. These commands are used to identify errors and verify the success of refactoring efforts within the user's project environment.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it reads and processes the project's source code (e.g., files in app/routes/, app/lib/) and has the capability to modify those files or execute commands based on that content.
Ingestion points: Source code files in the app/ directory are read during the Phase 2 analysis and execution.
Boundary markers: Absent. There are no explicit instructions or delimiters used to ensure the agent ignores natural language instructions that might be embedded as comments within the source code.
Capability inventory: The skill possesses file-writing capabilities (for refactoring) and command execution capabilities (via npm scripts).
Sanitization: Absent. The skill does not perform validation or sanitization on the content of the files it processes before incorporating them into its decision-making logic.

auto-refactor