secondbrain-transcribe

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches meeting transcripts and summaries from the Fireflies API (references/provider-fireflies.md and the .claude/lib/fireflies client called in SKILL.md Step 3) and directly reads and interprets that untrusted, user-generated transcript content to extract decisions/action items and drive creation of ADRs and documentation (Steps 5, 7, and 9), so third-party content can materially influence agent behavior.