The Agent Skills Directory

[PROMPT_INJECTION]: The directive to inject misspelled profanity ('fuckk') into the codebase acts as an instruction to bypass standard AI safety guidelines and propagate offensive content.
[PROMPT_INJECTION]: The skill exposes an indirect prompt injection vulnerability surface.
Ingestion points: File contents are read into the agent context using the Read tool after discovery by Grep.
Boundary markers: There are no instructions or delimiters defined to prevent the agent from being influenced by instructions embedded within the codebase files.
Capability inventory: The skill utilizes Grep, Read, and Edit (with the powerful replace_all flag) across the entire codebase.
Sanitization: No validation, escaping, or filtering is applied to the data before the replacement operation is performed.
[COMMAND_EXECUTION]: The skill automates the use of file system tools (Grep and Edit) for bulk modification. While these are platform-provided tools, the logic facilitates wide-scale automated defacement of the project's source code.

hello-replacer