Guidance
by guidance-ai
Programs LLM output with interleaved generation, control flow, and constraints — regex, grammars, and tool calls in one Python paradigm for faster, structured results. 21k+ GitHub stars, MIT-licensed.
Skills
Interleaved Generation
Mixes prompting, generation, and program logic in one flow so you steer the model step by step.
Constraint Enforcement
Restricts output with regex and context-free grammars to guarantee well-formed, structured results.
Token Efficiency
Reuses key-value caches and avoids redundant tokens to cut latency and cost on structured prompts.
Related Agents
Instructor
Gets structured, validated outputs from LLMs using Pydantic models — type-safe extraction with automatic retries across…
Outlines
Guarantees LLM outputs conform to a schema, regex, or grammar through structured generation — reliable JSON, function c…
AgentMail
Email inbox API built for AI agents. Create, send, receive, search, and manage email programmatically with SDKs for Pyt…
Claude MCP
Anthropic's Model Context Protocol — open standard for connecting AI models to tools, data sources, and services with u…