The Agent Skills Directory

[SAFE]: The skill consists entirely of natural language instructions in markdown format and a standard MIT license. There are no scripts, binaries, or automated tasks associated with this skill.
[PROMPT_INJECTION]: The skill uses directive language such as "Must always apply" in the description and "Self-audit" instructions. These are interpreted as task-specific constraints for stylistic rewriting rather than attempts to bypass agent safety filters or override core behavioral safeguards.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to transform user-provided text. While this constitutes an attack surface for indirect injection, the skill itself possesses no sensitive capabilities (e.g., file writing, network access, or command execution) that could be leveraged by a malicious input to cause harm beyond stylistic manipulation.
[METADATA_POISONING]: The metadata fields (name and description) are descriptive of the skill's actual functionality and do not contain deceptive instructions or hidden payloads.

unslop