senpi-smart-money

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow executes python3 scripts/smartmoney.py via the agent’s exec tool, where smartmoney.py reads outsider-authored free text indirectly from MCP tool responses (e.g., discovery_get_top_traders, discovery_get_trader_state, leaderboard_get_markets, leaderboard_get_momentum_events) and parses fields like trader positions and momentum events into the JSON that the LLM narrates.