Agent Tools¶
run_agent¶
Run a single AI analysis agent on a ticker.
ticker: "AAPL"
agent: "elvis_marlamov" # or warren_buffett, ben_graham, charlie_munger, cathie_wood, peter_lynch
Returns: signal (Bullish/Neutral/Bearish), score (0–100), confidence (0.0–1.0), rationale, breakdown by scoring blocks.
See Analysis Agents for details on each agent's methodology.
run_all_agents¶
Run all active agents on a ticker and compute consensus.
Returns: consensus score, signal, confidence, and per-agent results. Agents run concurrently via asyncio.gather.
run_recommendation¶
Generate a buy/hold recommendation with position sizing.
Returns: consensus, risk metrics, technical signals, position size (0–25% portfolio), stop-loss level.
Available Agents¶
| Agent | Style | Scoring Blocks |
|---|---|---|
elvis_marlamov |
Fundamentals + sentiment | valuation / quality / catalysts / financial_health |
warren_buffett |
Value investing | margin_of_safety / durable_advantage / management_quality |
ben_graham |
Deep value | net_net_value / earnings_stability / financial_strength |
charlie_munger |
Wonderful business at fair price | business_quality / fair_price / financial_fortress |
cathie_wood |
Innovation & growth | growth_signals / innovation_premium / market_position |
peter_lynch |
GARP | peg_value / earnings_quality / common_sense |
Each agent implements the AnalysisAgent protocol and returns an AgentResult with signal, score, confidence, and per-block breakdown.