Content blocked

Moderation runs on prompts and outputs. The current (MVP) check is a keyword filter on CSAM / explicit-minor / terror-instruction terms — the full safety stack (Vertex AI Safety + Google Cloud Vision SafeSearch + Hive fallback) lands in M5.

If a legitimate prompt was blocked, rephrase it and retry. If you believe the block is a false positive, reach out at safety@clipie.ai with the job id.