broll-finder

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

• Third-party content exposure detected (high risk: 0.85). Yes: the required runtime workflow has the agent read outsider-authored YouTube caption/transcript text from scripts/transcripts.py into found-broll/<slug>/transcripts/*.md at the “CHECKPOINT” in scripts/find_broll.py, which is then placed into the agent’s LLM context for window selection (indirect prompt injection risk).