wealth-career
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied birth data and external Qi Men chart transcriptions, which constitute an attack surface for indirect prompt injection.
- Ingestion points: User-provided birth information and transcribed chart data ingested via the
qimeninput field inlib/analyze.js. - Boundary markers: The skill implements strict interpretation boundaries, using
REPORT_CONTRACT.redlinesand explicit methodology constraints defined inmethodology.mdto prevent external data from overriding core logic. - Capability inventory: The skill's capabilities are limited to computational analysis and report generation. It does not perform network operations, file system writes, or shell command execution.
- Sanitization: Input data is sanitized through a normalization layer (
normalizeChart) that filters markers based on confirmation status and excludes raw source text (raw,source,school) from the final interpretation context.
Audit Metadata