The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external, untrusted text to perform humanization edits, which creates a surface for indirect prompt injection attacks where malicious instructions hidden in the text could override agent behavior. * Ingestion points: User-provided text files (referred to as FILE and PATH in SKILL.md and TESTING.md). * Boundary markers: None identified; the skill does not use delimiters or explicit instructions to treat the target content purely as data to be analyzed. * Capability inventory: The skill instructs the agent to use subprocess calls for wc, grep, awk, tr, and cat to analyze text metrics. * Sanitization: No input validation or content filtering is performed on the ingested source material.
[COMMAND_EXECUTION]: The skill employs standard system utilities including wc, grep, tr, and awk for quantitative text analysis, such as calculating word counts and measuring sentence length distribution. These commands are used locally for legitimate analytical purposes related to the skill's primary function.

humanize