turbosnap-debug-customer

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

• Third-party content exposure detected (high risk: 0.70). The skill instructs the agent to request and consume user-provided hosted metadata or pasted contents (e.g., .chromatic/preview-stats.trimmed.json or preview-stats.json) as part of its diagnostic workflow (see SKILL.md, reference/command-catalog.md and reference/evidence-ladder.md), so it will ingest untrusted third-party/user-generated content that can materially affect diagnosis and next actions.