geo-llmstxt

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

• Third-party content exposure detected (high risk: 1.00). This skill's Analysis and Generation workflows explicitly instruct the agent to use WebFetch to retrieve and crawl arbitrary public sites (e.g., "[domain]/llms.txt", "/llms-full.txt", homepage, sitemap.xml" and page contents) and to read H1/meta/paragraphs — clearly ingesting untrusted third-party web content that can influence subsequent parsing, validation, and generation decisions (SKILL.md "Analysis Mode" Step 1 and "Generation Mode" Step 1–3).

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches and ingests runtime files from a site's root (e.g., https://{domain}/llms.txt and https://{domain}/llms-full.txt) which are then parsed to guide the agent's behavior and outputs, so these remote files directly control prompts/instructions at runtime.