firecrawl-monitoring

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md describes a Firecrawl webhook that “becomes a release” and an ingest workflow that parses diff text and uses AI prompts for judging/selection; since Firecrawl is an external fetch of arbitrary web pages, the webhook payload can include outsider-authored page content/diffs that are ingested into the agent/LLM context at runtime.