from aiskillstore
Rewrite AI-generated text into more natural, human-like prose using the HumanizerAI API, with before/after scores and credit tracking.
Transforms AI-generated content into natural, human-like writing by calling the HumanizerAI API. The skill accepts text and an optional intensity flag (light, medium, aggressive) and returns the humanized output alongside scoring and credit usage details.
Invoke when users want to reduce detectability of AI-generated text, improve naturalness, or adjust rewriting intensity. Typical triggers include commands like /humanize or requests mentioning AI detectors, bypassing detectors, or needing a more human tone.
Agents that can call external web APIs (WebFetch/HTTP) and handle argument parsing; suitable for Claude, Codex, or any agent with web access and token management.
This skill instructs an agent to call the HumanizerAI API to rewrite AI-generated text as more human-like prose. The SKILL.md provides clear API call details, response format, and error handling guidance. No scripts are bundled — it's purely an instruction-only skill that relies on the agent having a HUMANIZERAI_API_KEY. The skill encourages bypassing AI detection tools like Turnitin and GPTZero, which raises ethical concerns, though no direct security vulnerabilities exist in the instructions themselves.
The skill's primary purpose is helping users evade AI detection systems, which is ethically questionable especially regarding academic tools like Turnitin. However, from a pure security standpoint, the instructions are straightforward API call guidance with no malicious code, exfiltration, or destructive commands. The architecture is thin — no scripts, no references/, just a single SKILL.md with API instructions. Frontmatter is present and valid. Code quality is moderate — instructions are clear enough but lack depth for edge cases.
