run_decision_pipeline
Run a multi-stage decision pipeline that combines counterfactual analysis, scenario generation, and foresight planning to evaluate high-stakes objectives and deliver ranked recommendations with evidence.
Instructions
Run a multi-stage decision pipeline with configurable reasoning stages.
Orchestrates counterfactual reasoning, scenario generation, foresight planning, meta-reasoning, uncertainty estimation, and information-need detection into a single pipeline run. The pipeline evaluates the provided objective and returns a ranked recommendation with evidence.
Use this for high-stakes decisions where multiple perspectives add value. For simpler individual analyses, use the standalone tools (evaluate_plan, generate_scenarios, etc.) instead.
Side effects: Records decision events; if execute_mode='queued_runtime', may enqueue follow-up tasks for runtime execution.
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | Maximum events to process (default 10000). | |
| objective | Yes | The decision objective dict describing what to decide. | |
| max_horizon | No | Max planning depth (default 5). | |
| execute_mode | No | 'event_only' (record only), 'queued_runtime' (queue tasks), 'simulate', or 'validate' (default 'event_only'). | event_only |
| risk_threshold | No | Maximum acceptable risk score 0-1 (default 0.7). | |
| foresight_limit | No | Max action chains per branch (default 5). | |
| scenarios_limit | No | Max scenarios to generate (default 4). | |
| enable_foresight | No | Run foresight planning (default False). | |
| enable_scenarios | No | Run scenario generation (default False). | |
| regret_threshold | No | Regret threshold 0-1 for counterfactual filtering (default 0.2). | |
| enable_uncertainty | No | Estimate epistemic/aleatoric uncertainty (default False). | |
| counterfactual_limit | No | Max alternative outcomes to generate (default 3). | |
| enable_counterfactual | No | Run counterfactual what-if analysis (default False). | |
| enable_meta_reasoning | No | Apply meta-reasoning over results (default False). | |
| simulate_before_execute | No | Simulate effects before committing (default False). | |
| enable_information_seeking | No | Identify information gaps (default False). | |
| scenario_recommendation_threshold | No | Min score 0-1 for scenario inclusion (default 0.5). |
Output Schema
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||