chain_of_draft_solve
Solve reasoning problems efficiently using the Chain of Draft approach, minimizing token usage while maintaining accuracy. Specify problem, domain, and adaptive word limits for precise solutions.
Instructions
Solve a reasoning problem using Chain of Draft approach
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| adaptive_word_limit | No | Adjust word limits based on complexity | |
| approach | No | Force 'CoD' or 'CoT' approach | |
| domain | No | Domain for context (math, logic, code, common-sense, etc.) | |
| enforce_format | No | Whether to enforce the word limit | |
| max_words_per_step | No | Maximum words per reasoning step | |
| problem | Yes | The problem to solve |
Input Schema (JSON Schema)
{
"properties": {
"adaptive_word_limit": {
"description": "Adjust word limits based on complexity",
"type": "boolean"
},
"approach": {
"description": "Force 'CoD' or 'CoT' approach",
"type": "string"
},
"domain": {
"description": "Domain for context (math, logic, code, common-sense, etc.)",
"type": "string"
},
"enforce_format": {
"description": "Whether to enforce the word limit",
"type": "boolean"
},
"max_words_per_step": {
"description": "Maximum words per reasoning step",
"type": "number"
},
"problem": {
"description": "The problem to solve",
"type": "string"
}
},
"required": [
"problem"
],
"type": "object"
}