trade-sentiment
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill performs standard financial research tasks using authorized web search tools to gather publicly available information on market sentiment, news, and institutional filings.
- [SAFE]: The skill writes analysis results to a local file in the current working directory, which is expected behavior for an automated reporting tool.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external web search results (news headlines, social media posts from Reddit and X, and analyst reports).
- Ingestion points: Data enters the agent context through the
WebSearchtool as specified in the 'Data Gathering' section of SKILL.md. - Boundary markers: The prompt does not provide specific delimiters or instructions to ignore potential commands embedded in search results.
- Capability inventory: The agent is restricted to web search and file writing (markdown format).
- Sanitization: The agent is instructed to summarize and score the data, which involves transformation but no explicit sanitization of the raw input text.
Audit Metadata