llmframe 2.1.0

OpenAI-first hexagonal Python application scaffold for llmframe.

llmframe

OpenAI-first Python hexagonal application scaffold for the llmframe repository.

Requirements

• Python 3.11+
• uv for environment and dependency management

Setup

Install the project and development dependencies:

uv sync --all-extras

Project layout

The repository follows a hexagonal architecture layout:

src/llmframe/
├── domain/          # Pure business logic and invariants
├── application/     # Use cases and port interfaces
│   └── ports/
└── adapters/        # Input/output adapters at the system boundary
    ├── input/
    └── output/
tests/
├── unit/
└── integration/

LLM adapters

The repository includes reusable LLM output adapters under llmframe.adapters.output.llm. Today this package is intentionally OpenAI-first: OpenAI is the only implemented provider integration, while the surrounding structure is being kept hexagonal so additional providers can be added later without leaking provider-specific concerns into the shared/application layers.

Key package areas:

• llmframe.adapters.output.llm.llm_adapter - provider-neutral high-level structured JSON and text generation adapter
• llmframe.adapters.output.llm.providers.openai - OpenAI provider adapter, client builder, transport, DTOs, and parsing helpers
• llmframe.adapters.output.llm.usage_tracker - aggregated token/cost tracking utilities

The sibling providers/anthropic and providers/gemini directories are architectural placeholders only; they are not wired integrations yet.

Example imports:

from llmframe.adapters.output.llm import LlmAdapter
from llmframe.adapters.output.llm.providers.openai import OpenAIClientSettings, build_provider
from llmframe.adapters.output.llm.usage_tracker import LlmUsageTrackerConfig, OpenAILlmUsageTracker

Quality checks

Run the local quality gate with uv:

uv run ruff format .
uv run ruff check .
uv run mypy .
uv run pytest

Release publishing

GitHub Actions publishes this package to PyPI using the workflow at .github/workflows/ci_cd.yaml.

• Publishing is triggered by pushes to the master branch or manual workflow dispatch.
• The workflow uses Python Semantic Release to determine the next version from conventional commits, create the release commit and tag, and generate changelog updates.
• pyproject.toml version updates are release-managed: do not bump project.version in regular feature or fix commits; Semantic Release writes the new version during the release commit.
• Package artifacts are built in CI and published to PyPI only.
• If no releasable commits are detected, the publish job is skipped.
• Publishing targets the llmframe project on pypi.org.

Before the workflow can publish successfully, configure a trusted publisher in the PyPI project settings for this GitHub repository and workflow file.

Also ensure the workflow has permission to use the repository GITHUB_TOKEN for creating release commits and tags.

Example release flow:

git commit -m "feat: add new llm transport option"
git push origin master

Notes

• The import package name is llmframe.
• Keep business logic inside domain/ and application/.
• Keep framework, transport, persistence, and integration concerns in adapters/.