ln-612-semantic-content-auditor

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs the worker to fetch missing shared files via WebFetch from a public GitHub raw URL (https://raw.githubusercontent.com/levnikolaevich/claude-code-skills/master/skills/{path}) and lists multiple "MANDATORY READ" reference files the agent must load and interpret as part of its workflow, meaning public, third‑party content can be ingested and materially influence decisions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly instructs at runtime to fetch required "MANDATORY READ" contract files from https://raw.githubusercontent.com/levnikolaevich/claude-code-skills/master/skills/{path}, and those fetched documents (audit contracts/templates) directly control the auditor's prompts and behavior, making this an external runtime dependency that can alter agent instructions.