parsing-changelogs

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md describes runtime fetching of changelog content from external URLs/feeds (e.g., “Feed adapter” / “Markdown fetch” / “Fast fetch” / “Cloudflare rendering” / “Firecrawl monitoring”), which ingests outsider-authored free text (page/HTML/markdown content) into the parsing pipeline and thus into the agent’s LLM context.