The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes Grep and Edit tools to perform read and write operations on the local filesystem. It also describes a workflow for executing Python scripts via a separate 'code-execution' skill for bulk refactoring operations.
[DYNAMIC_EXECUTION]: For batches involving 10 or more files, the skill automatically switches to a code execution mode. It provides examples of generating and running Python scripts using a local module (api.code_transform) to perform bulk modifications, which constitutes dynamic script execution.
[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from external source code files and uses that data to drive file modification tasks.
Ingestion points: The skill reads local source files using the Grep and Edit tools as part of its core refactoring workflow.
Boundary markers: The instructions do not provide explicit boundary markers or specific guidance to the agent to ignore or isolate instructions that may be embedded within the code being processed.
Capability inventory: The skill possesses the capability to read from and write to the local filesystem and explicitly leverages code execution for large-scale tasks.
Sanitization: There is no documentation or implementation of sanitization, validation, or escaping for the content ingested from the external files before it is processed by the agent.

code-refactor