oma-recap

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill ingests and outputs raw conversation prompts (via oma recap --json or the jq fallback which pulls .display text) and saves/displayes them in recaps without any redaction rules, so any secrets present in histories could be reproduced verbatim in the generated output.