prediction-market-strategy

Pass

Audited by Gen Agent Trust Hub on Jun 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists of strategy documentation and a single utility script (scripts/sizing.py). All logic is transparent and matches the stated purpose of sizing and backtesting strategies for prediction markets.
  • [COMMAND_EXECUTION]: There is no command execution or subprocess spawning. The provided Python script is a pure-math utility using only the standard math library.
  • [DATA_EXFILTRATION]: No network access or data exfiltration patterns were detected. The scripts do not perform any HTTP requests or access the local file system for sensitive data.
  • [CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or credentials were found. The documentation refers to external API tools (e.g., kalshi-api) but does not implement them or require secrets here.
  • [PROMPT_INJECTION]: No prompt injection attempts, instructions to bypass safety guidelines, or role-play markers were detected in the instructions or metadata.
  • [EXTERNAL_DOWNLOADS]: The skill does not perform any external downloads or package installations. It relies on the Python standard library.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 24, 2026, 08:10 PM
Security Audit — agent-trust-hub — prediction-market-strategy