The Agent Skills Directory

[SAFE]: The skill follows secure patterns by restricting tool usage to the provided manifest and implementing strict output constraints regarding financial advice. The use of a configuration file for sentiment thresholds is a standard and safe software practice.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it processes external market data via various tool calls. However, this is considered low risk given the specific context of financial data and the total absence of sensitive sinks such as file system modification, command execution, or arbitrary network exfiltration. 1. Ingestion points: Market data tools listed in SKILL.md. 2. Boundary markers: Absent from the prompt instructions. 3. Capability inventory: The skill is restricted to generating text analysis and does not possess tools for persistent storage or external communication. 4. Sanitization: No explicit data sanitization is performed on the tool outputs before analysis.

market-scanner