The Agent Skills Directory

[COMMAND_EXECUTION] (LOW): Multiple analyst sub-skills (fundamentals-analyst.md, market-analyst.md, news-analyst.md, social-analyst.md) use curl to interact with the Alpha Vantage API. While the use case is legitimate, the direct execution of network commands with variable-injected parameters creates a minor attack surface if input ticker symbols are not properly validated by the calling agent.
[EXTERNAL_DOWNLOADS] (SAFE): The skill downloads financial data from Alpha Vantage. This is a well-known, trusted source for financial market data, and the data is restricted to CSV/JSON formats for analysis rather than executable code.
[INDIRECT_PROMPT_INJECTION] (LOW): The skill is vulnerable to indirect prompt injection because it processes untrusted data from the open web.
Ingestion points: External news articles and social media sentiment are ingested via the NEWS_SENTIMENT function in news-analyst.md and social-analyst.md.
Boundary markers: Absent. There are no instructions or delimiters provided to help the agent distinguish between its own logic and instructions that might be embedded within the news text.
Capability inventory: The skill possesses the ability to perform filesystem writes (creating reports) and execute network requests via curl.
Sanitization: Absent. There is no evidence of filtering or escaping logic to prevent malicious payloads in news headlines from influencing the 'Risk Manager' or 'Trader' agents later in the pipeline.

trading-agents