write-gitbook

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow can fetch GitBook documentation pages via an HTTP GET to https://gitbook.com/docs/skill.md?ask=..., which returns excerpts from web content that may include outsider-authored free text, and that response is then ingested into the agent’s LLM context.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs the agent at runtime to GET https://gitbook.com/docs/skill.md?ask= and use the returned text as direct answers/excerpts injected into the agent's context, meaning remote content can control the agent's prompts/responses.