The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests and processes untrusted data (text to humanize and voice calibration samples) without adequate security boundaries.
Ingestion points: SKILL.md defines workflows for humanizing provided text and reading voice samples from user-specified file paths.
Boundary markers: The instructions fail to define delimiters or safety constraints to prevent the agent from executing instructions contained within the input data.
Capability inventory: The skill is configured with access to Read, Write, Edit, Grep, and Glob tools as listed in the SKILL.md frontmatter.
Sanitization: The skill does not implement or require any form of input validation or content filtering for the processed text.

humanizer