url-extractor

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill explicitly extracts and outputs full URLs and uses a fetch tool to access them, which can force the LLM to reproduce any embedded sensitive tokens or credentials (e.g., API keys in query strings) verbatim in its output.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). In the skill’s workflow, the agent first extracts URLs from user-provided free text and then uses the web_fetch tool to access each extracted URL and extract page titles, meaning outsider-authored text can drive LLM ingestion of arbitrary remote page content at runtime.