added yaml to the requiements so the tests won't fail

AI fixes tests

fixed CI tests

AI eventually fixed tests and completions

AI tries to fix tests v4

AI tries to fix tests v3

AI tries to fix it's own tests v2

AI fixed generated tests

AI added tests

Implement Responses API

Use live API for OpenAI responses client test

added GPT4-dev-177M-1511-Instruct

cosmetic / changed url from localhost to HF space

A cosmetic push to check the workflows again

fix the error with doubleing "sources field"

finally added the new model in hopes this will work

Fix list models endpoint to exclude metadata

Fix test_model_detail_serialization by removing metadata from list endpoint

Initial plan

debug model addition to registry#3

Merge branch 'main' into codex/debug-model-addition-to-registry

Return metadata only in model detail endpoint

debug model addition to registry#2

Merge branch 'main' into codex/debug-model-addition-to-registry

Include metadata in model listings

debug model addition to registry

Merge branch 'main' into codex/debug-model-addition-to-registry

Add GPT4-dev model to default registry

Merge pull request #9 from krll-corp/codex/debug-model-addition-to-registry

Revert "Add GPT4-dev-177M-1511 model specification"

Merge pull request #8 from krll-corp/codex/debug-model-addition-to-registry

Merge branch 'main' into codex/debug-model-addition-to-registry

Load defaults only when explicitly requested with custom registry

Add regression coverage for /v1/models response

Add regression test for /v1/models

Allow combining default registry with custom files

Allow extending default model registry

added a direct HF space link

Add GPT4-dev-177M-1511 model specification

Fail live model probe on API errors

Exercise live models against Hugging Face API

Fail live model probe on API errors

Add tests workflow

Parallelize CI tests across envs

Add support for batched prompts + advanced model metadata

Support prompt batching and enrich model metadata

actualized README

removed an extra space between the prompt and a completion

Added more advanced health monitoring

Add endpoint health monitoring to root status