instructor · Issues· 66 open

Remote media type detection fails when servers reject HEAD requests

SSRF filter bypass: is_global misses IPv6 transition addresses (NAT64/IPv4-compat) → cloud metadata

Handle non-openai conform List[contentitem] responses

context= templating is a silent no-op on the Bedrock provider

Gemini structured-output reask always fails: request ends on a model turn

Public `retry_sync`/`retry_async`/`handle_reask_kwargs` mutate the caller's `messages` list during validation retries

Feature request: OWASP ASI06 memory poisoning defense for structured output agents

`is_simple_type` falsely returns True for custom non-Pydantic classes on Python 3.10+, causing schema generation crashes

`prepare_response_model` creates a new Pydantic class on every request, leaking one model class + validator per call

feat(bedrock): support Amazon Bedrock API key (bearer token) authentication

feat: Add persistent memory cookbook with Dakera integration

Feature: Memory-backed extraction — recall past structured outputs for few-shot context

Add Azure AI Foundry as a first-class v2 provider

Token-efficient serialization to reduce structured output overhead in production pipelines

feat: add MiniMax provider support