Ultrabrain Think
ultrabrain_thinkPerforms step-by-step reasoning for code engineering, enabling branching, revisions, quality checks, and bias detection within a structured thought chain.
Instructions
Canonical LCV reasoning gate for code work, branching, revisions, quality metrics, bias checks, confidence, and meta checkpoints.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| session_id | No | Optional reasoning session id. Defaults to "default". | |
| response_format | No | Response format. | |
| thought | Yes | Current Ultrabrain reasoning step. | |
| thought_number | Yes | Current thought number in the chain. | |
| total_thoughts | Yes | Estimated total thoughts. Adjust this as scope changes. | |
| next_thought_needed | Yes | Set to false only when this chain has reached a verified conclusion. | |
| step_type | No | Reasoning step category. | |
| mode | No | Reasoning mode. | |
| is_revision | No | Whether this step revises an earlier thought. | |
| revises_thought | No | Thought number being revised. | |
| branch_from_thought | No | Thought number where this branch starts. | |
| branch_id | No | Branch identifier. | |
| parent_thought | No | Optional parent thought reference. | |
| needs_more_thoughts | No | Allows thought_number to exceed total_thoughts when scope expands. | |
| depth_level | No | Current depth for serial reasoning. | |
| max_depth | No | Maximum planned depth. | |
| budget_mode | No | Reasoning budget mode. | |
| budget_used | No | Budget used percentage from 0 to 100. | |
| confidence | No | Confidence from 0 to 1. | |
| meta_checkpoint | No | Marks an explicit meta-reasoning checkpoint. | |
| bias_detected | No | Known cognitive bias to track. | |
| quality_metrics | No | Quality scores from 0 to 5. | |
| evidence | No | Evidence supporting this thought. | |
| assumptions | No | Assumptions to track. | |
| open_questions | No | Unresolved questions. | |
| alternatives | No | Alternative paths or options. | |
| risks | No | Known risks. | |
| next_actions | No | Concrete next checks or implementation actions. | |
| tags | No | Optional tags. | |
| perspective | No | Optional perspective, such as reviewer, maintainer, security, UX, or operator. | |
| expected_output | No | Expected output from the reasoning chain. | |
| hypothesis | No | Explicit hypothesis for this step. | |
| verification | No | Verification approach or result. |