The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill reads and processes source code from the repository to perform its audit (Step 2). This content is untrusted and could contain malicious instructions hidden in comments or strings designed to influence the agent's behavior during the audit or refactoring process.
Ingestion points: Source files changed in the diff are read in full using the Read tool.
Boundary markers: No explicit delimiters or instructions to ignore embedded prompts are defined for the files being read.
Capability inventory: The skill uses Bash to execute commands and the Write and Edit tools to modify project files.
Sanitization: No sanitization or filtering of the file content is specified before processing.
[COMMAND_EXECUTION]: The skill is designed to find and execute test commands defined in the project's configuration files (e.g., package.json, Makefile, pyproject.toml). This results in the execution of workspace-provided shell commands, which is necessary for its function but represents a point of execution for arbitrary repository code.

clean-code