analyzing-market-sentiment

Warn

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: MEDIUMCREDENTIALS_UNSAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The documentation in references/implementation.md directs the agent to use the Read tool to access sensitive API credentials from a configuration file at {baseDir}/config/crypto-apis.env.
  • [COMMAND_EXECUTION]: The script scripts/news_sentiment.py implements dynamic loading of code by modifying sys.path at runtime to import modules from a computed path relative to the script location.
  • [EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch data from well-known external sources including the Alternative.me Fear & Greed API, the CoinGecko API, and various news RSS feeds (CoinTelegraph, CoinDesk, Decrypt).
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection. Ingestion points: Untrusted news headlines and summaries fetched in scripts/news_sentiment.py. Boundary markers: Absent in tool output. Capability inventory: Execution of Python scripts via the Bash tool. Sanitization: None provided for the raw news text presented to the agent for interpretation.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 13, 2026, 06:59 AM