The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection (Category 8) as it processes untrusted data from local source code files.
Ingestion points: The skill reads source files using Read, Grep, and Glob tools to analyze code structure and imports.
Boundary markers: There are no explicit instructions to the agent to ignore embedded instructions or comments within the processed files.
Capability inventory: The skill utilizes Bash to run tests and perform file system operations (e.g., find, sed for auto-fixing).
Sanitization: No sanitization or validation of the content read from files is performed before processing.
[COMMAND_EXECUTION]: The skill includes instructions for executing shell commands to modify or delete files.
Evidence: The 'Auto Repair' section includes sed for batch string replacement in source files and find with the -delete flag to remove empty files.
Mitigation: The YAML frontmatter attempts to restrict Bash usage to specific test and compile commands (npm test:*, yarn test:*, npx tsc:*), which limits the execution surface compared to the examples in the documentation.

refactor-check