financial-analysis

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt requires parsing and embedding user-provided vault files (e.g., bank statements, account files, Kontoauszüge) into generated reports without any redaction rules, so the LLM may be required to output sensitive financial identifiers or secret values verbatim.

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow ingests outsider-authored free text via finance/docs/ transcriptions (e.g., “Standmitteilungen, Kontoauszuge, Arbeitsvertrag, Nebenkostenabrechnungen”), which are documents the operating user did not author and whose German text is parsed into the LLM context.