The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various development tools via the shell to perform its tasks.
Runs git commands for churning, ownership, and hotspot analysis.
Invokes test runners like vitest, pytest, and flutter_test to verify behavior preservation.
Uses uvx to run metrics tools such as lizard and radon for complexity analysis.
[INDIRECT_PROMPT_INJECTION]: The skill analyzes and modifies user-provided source code, which serves as an untrusted ingestion point.
Ingestion points: Target source files and modules specified by the user in the target parameter.
Boundary markers: No explicit boundaries or warnings are implemented to differentiate code content from agent instructions.
Capability inventory: The skill has the ability to write to the file system and execute shell commands (git, test runners, analysis tools).
Sanitization: No specific sanitization or filtering of source code content is performed.
[DYNAMIC_EXECUTION]: Uses uvx to dynamically fetch and execute code analysis tools.
Fetches and runs lizard and radon packages from the standard Python registry.
Mentions the use of transformation engines like jscodeshift and ast-grep for automated refactoring.

oma-refactor