Gpt Analyzer
PassAudited by ClawScan on May 1, 2026.
Overview
This is a benign text-analysis skill with no system access, though its GPT-detection confidence should not be treated as definitive.
This skill appears safe to install from an access-control perspective because it does not request system, account, network, or file privileges. Treat its GPT/model detection output as a rough indicator rather than proof.
Findings (1)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Users could over-trust the result and misclassify human or AI-written text if they treat the confidence score as definitive.
The skill presents model fingerprinting/version identification, while the shown implementation relies on simple phrase, list-structure, and sentence-uniformity heuristics.
description: GPT-specific pattern detection with model fingerprinting and version identification
Use this only as a heuristic signal; verify important decisions with additional evidence and disclose limitations around false positives and false negatives.