finance-research-agent
Pass
Audited by Gen Agent Trust Hub on Feb 28, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill retrieves market data and financial information from trusted and well-known technology providers. Evidence includes the use of Yahoo Finance via the yfinance library, CoinGecko via pycoingecko, and the Federal Reserve Economic Data via the fredapi. These sources are considered safe and are appropriate for the skill's intended financial research purpose.
- [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection through its ingestion of external news content. 1. Ingestion points: Untrusted headlines and snippets are fetched from external RSS feeds in scripts/finance_api.py. 2. Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the Jinja2 templates found in assets/templates/. 3. Capability inventory: The agent has the ability to perform network requests to financial APIs and write files to the local workspace. 4. Sanitization: News content is directly interpolated into markdown reports without validation or escaping, allowing potentially malicious instructions in headlines to influence the agent's behavior.
Audit Metadata