The Agent Skills Directory

[COMMAND_EXECUTION]: The skill generates shell commands using the user-provided asset name to search the codebase. This approach presents a risk of command argument injection if the input name is not correctly handled during command construction.
[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection because it ingests and processes the content of numerous project files to determine removal impacts.
Ingestion points: Data from grep results across agents/, rules/, skills/, README.md, and hooks/hooks.json (SKILL.md, references/removal-policy.md).
Boundary markers: The instructions lack explicit delimiters or safety warnings to ensure the agent ignores instructions embedded within the scanned file contents.
Capability inventory: The skill is configured with Edit, Write, Bash (file deletion), and AskUserQuestion (SKILL.md).
Sanitization: There is no specified sanitization or filtering logic for the data extracted from the project files before it is used to influence the removal plan.

safe-remove