post-market-debrief
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists entirely of natural language instructions and a structured output template. It does not contain any executable scripts, binaries, or system configurations that could be used for malicious purposes.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external financial data, creating a potential surface for indirect prompt injection. However, the risk is categorized as safe because the skill's capabilities are restricted to text generation within a fixed reporting structure. 1. Ingestion points: Market data fetched via the wind.financial.data tool and user-provided observations. 2. Boundary markers: Absent; the prompt does not use specific delimiters to isolate external data. 3. Capability inventory: The agent is limited to analysis and text generation; no file system, shell, or network exfiltration capabilities are present. 4. Sanitization: None; the skill relies on the underlying model's inherent safety filters to handle potentially adversarial content in the ingested data.
Audit Metadata