What problem does it solve? LLMs frequently produce malformed JSON, invalid emails, or off-format outputs that break downstream parsing and require retry loops. This Skill uses Microsoft Research's Guidance framework to enforce regex and grammar constraints at the token level, so generated output always matches the required structure. ## Core Features & Use Cases - Regex and Grammar Constraints: Guarantee valid emails, dates, phone numbers, JSON objects, and code identifiers by filtering invalid tokens during generation. - Token Healing: Automatically repairs token boundaries between prompt and generation to avoid spacing artifacts and improve output quality. - Multi-Step Workflows and Agents: Build ReAct agents, chain-of-thought reasoning, and stateful generation functions with the @guidance decorator. - Use Case: Extract structured entities (person, organization, date, location) from unstructured text with guaranteed date formats and valid field values, ready for direct database insertion. ## Quick Start Ask the agent to use the Guidance library to generate a JSON user profile with constrained name, age, and email fields using an Anthropic or OpenAI model.
