Extract structured data from LLM responses with Pydantic validation, retry failed extractions automatically, parse complex JSON with type safety, and stream partial results with Instructor - battle-tested structured output library
This skill does not declare a tool allowlist. The agent host applies whatever default tools are available at runtime.
SKILL.md / Manifest
https://raw.githubusercontent.com/zechenzhangagi/ai-research-skills/main/16-prompt-engineering/instructor/SKILL.mdRegistry
github (via claudemarketplaces.com)