guidance - Constrain LLM generation with Guidance
Controls LLM output with regex and grammars to generate valid structured data and build multi-step workflows.
Tags
Updated: 2026-09-29Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Control output with regex
- Generate valid structured data
- Enforce grammar constraints
- Select from fixed options
- Heal token boundaries
- Build multi-step workflows
- Integrate model backends
Inputs
- LLM backend
- API credentials
- Generation prompts
- Regex patterns
- Grammar definitions
- Selection options
- Tool functions
Outputs
- Constrained text
- Structured JSON, XML, or code
- Named generation fields
- Workflow results
Requirements
- Python environment
- guidance package
- transformers package
- Supported model backend
- Backend credentials or local model