build(deps-dev): Update langchain-openai requirement from <1,>=0.3.0 to >=0.3.0,<2

.github/CODEOWNERS

@@ -0,0 +1 @@
+* @gritaro

.github/dependabot.yaml

@@ -0,0 +1,17 @@
+# To get started with Dependabot version updates, you'll need to specify which
+# package ecosystems to update and where the package manifests are located.
+# Please see the documentation for all configuration options:
+# https://docs.github.com/github/administering-a-repository/configuration-options-for-dependency-updates
+
+version: 2
+updates:
+- package-ecosystem: "pip"
+  directory: "/"
+  schedule:
+    interval: "daily"
+
+- package-ecosystem: "github-actions"
+  directory: "/"
+  schedule:
+    # Check for updates to GitHub Actions every weekday
+    interval: "daily"

.github/settings.yml

@@ -0,0 +1,36 @@
+# These settings are synced to GitHub by https://probot.github.io/apps/settings/
+
+repository:
+  name: ha-smartchain
+  description: "SmartChain — multi-provider LLM conversation agent for Home Assistant (GigaChat, YandexGPT, OpenAI) with tool calling and streaming"
+  homepage: https://github.com/gritaro/ha-smartchain
+  topics: homeassistant, hacs-integration, voice-assistant, langchain, gigachat, yandexgpt, openai, llm, smart-home, ai-agent
+  private: false
+  has_issues: true
+  has_projects: false
+  has_wiki: false
+  default_branch: main
+  allow_squash_merge: true
+  use_squash_pr_title_as_default: true
+  allow_merge_commit: false
+  allow_rebase_merge: true
+  delete_branch_on_merge: true
+  enable_vulnerability_alerts: true
+
+labels:
+- name: "Feature Request"
+  color: "00ffbb"
+- name: "Bug"
+  color: "e30000"
+- name: "Wont Fix"
+  color: "ffffff"
+- name: "Enhancement"
+  color: "48ff00"
+- name: "Documentation"
+  color: "0077ff"
+- name: "Help needed"
+  color: "fbca04"
+- name: "dependencies"
+  color: "000000"
+- name: "github_actions"
+  color: "000000"

.github/workflows/cron.yaml

@@ -0,0 +1,21 @@
+name: Cron HACS actions
+
+on:
+  schedule:
+  - cron: "0 0 * * *"
+
+jobs:
+  validate:
+    runs-on: "ubuntu-latest"
+    name: Validate
+    steps:
+    - uses: "actions/checkout@v4"
+
+    - name: HACS validation
+      uses: "hacs/action@main"
+      with:
+        category: "integration"
+        ignore: brands
+
+    - name: Hassfest validation
+      uses: "home-assistant/actions/hassfest@master"

.github/workflows/pull.yml

@@ -0,0 +1,40 @@
+name: Pull actions
+
+on:
+  pull_request:
+
+env:
+  PYTHON_VERSION: "3.12"
+
+jobs:
+  validate:
+    runs-on: "ubuntu-latest"
+    name: Validate
+    steps:
+    - uses: "actions/checkout@v4"
+
+    - name: HACS validation
+      uses: "hacs/action@main"
+      with:
+        category: "integration"
+
+    - name: Hassfest validation
+      uses: "home-assistant/actions/hassfest@master"
+
+  lint:
+    runs-on: "ubuntu-latest"
+    name: Lint with ruff
+    steps:
+    - uses: "actions/checkout@v4"
+    - uses: "astral-sh/setup-uv@v5"
+    - run: uv run ruff check .
+    - run: uv run ruff format --check .
+
+  test:
+    runs-on: "ubuntu-latest"
+    name: Run tests
+    steps:
+    - uses: "actions/checkout@v4"
+    - uses: "astral-sh/setup-uv@v5"
+    - run: uv sync
+    - run: uv run pytest tests/ -v

.github/workflows/push.yml

@@ -0,0 +1,44 @@
+name: Push actions
+
+on:
+  push:
+    branches:
+    - main
+    - rc-*
+
+env:
+  PYTHON_VERSION: "3.12"
+
+jobs:
+  validate:
+    runs-on: "ubuntu-latest"
+    name: Validate
+    steps:
+    - uses: "actions/checkout@v4"
+
+    - name: HACS validation
+      uses: "hacs/action@main"
+      with:
+        category: "integration"
+        ignore: brands
+
+    - name: Hassfest validation
+      uses: "home-assistant/actions/hassfest@master"
+
+  lint:
+    runs-on: "ubuntu-latest"
+    name: Lint with ruff
+    steps:
+    - uses: "actions/checkout@v4"
+    - uses: "astral-sh/setup-uv@v5"
+    - run: uv run ruff check .
+    - run: uv run ruff format --check .
+
+  test:
+    runs-on: "ubuntu-latest"
+    name: Run tests
+    steps:
+    - uses: "actions/checkout@v4"
+    - uses: "astral-sh/setup-uv@v5"
+    - run: uv sync
+    - run: uv run pytest tests/ -v

.gitignore

@@ -0,0 +1,14 @@
+__pycache__
+.idea
+/shelf/
+/workspace.xml
+/httpRequests/
+/dataSources/
+/dataSources.local.xml
+*.iws
+*.iml
+*.ipr
+/.mcp.json
+/.serena/
+.pytest_cache
+/secrets/

.pre-commit-config.yaml

@@ -0,0 +1,6 @@
+repos:
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: v0.9.7
+    hooks:
+      - id: ruff
+      - id: ruff-format

CHANGELOG.md

@@ -0,0 +1,115 @@
+# Changelog
+
+All notable changes to this project are documented in this file.
+
+Format based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
+project follows [Semantic Versioning](https://semver.org/).
+
+## [0.9.0] - 2026-03-10
+
+### Added
+- **Sub-entries** — multiple conversation agents per provider via `ConfigSubentryFlow`. Each sub-entry has its own model, prompt, temperature, LLM API, and creates independent ConversationEntity + AITaskEntity
+- `ConversationSubentryFlow` with `async_step_user()` and `async_step_reconfigure()` for adding/editing agents
+- `async_get_supported_subentry_types()` on ConfigFlow returns `{"conversation": ConversationSubentryFlow}`
+- Backward compatible: entries without sub-entries continue working in legacy mode (single agent from `entry.options`)
+- **Options Flow tests** — 7 new tests covering form display, model validation, GigaChat-specific fields, LLM API handling
+- **E2E tool calling loop test** — full simulation: user request → tool_call → tool execution → final response
+- **Sub-entries tests** — 8 tests covering subentry flow, setup with subentries, multiple agents, legacy fallback
+
+### Changed
+- `conversation.py` — `SmartChainConversationEntity` now accepts `subentry_id` and `options` params; uses `_agent_options` and `_client` properties
+- `ai_task.py` — `SmartChainAITaskEntity` now accepts `subentry_id`; uses `_client` property
+- `__init__.py` — `async_setup_entry` creates per-subentry clients dict or single legacy client
+- `config_flow.py` — renamed `common_config_option_schema()` → `_subentry_schema()` (backward-compatible alias kept)
+- `OptionsFlow` — removed `__init__` (config_entry is read-only property in modern HA)
+- Total: 67 tests passing
+
+### Fixed
+- `OptionsFlow.__init__` — removed setter for `self.config_entry` (read-only property in HA 2025+)
+
+## [0.8.0] - 2026-03-10
+
+### Added
+- **Ollama provider** — local models (Llama 3.3, Qwen3, Gemma 3, T-Pro 2, T-Lite, DeepSeek R1, Phi 4, Home-3B). Config: base_url only, no API key needed
+- **DeepSeek provider** — cheapest cloud LLM (deepseek-chat, deepseek-reasoner). Uses ChatOpenAI with DeepSeek base URL
+- **Anthropic provider** — Claude models (claude-sonnet-4-6, claude-haiku-4-5, claude-opus-4-6) via ChatAnthropic
+
+### Changed
+- **ai_task made optional** — no longer in manifest.json dependencies, detected dynamically. Fixes hang on HA versions without ai_task support
+- Config Flow extended with `async_step_ollama`, `async_step_deepseek`, `async_step_anthropic`
+- `client_util.py` — new client factories for Ollama (ChatOllama), DeepSeek (ChatOpenAI), Anthropic (ChatAnthropic)
+- `manifest.json` — added `langchain-anthropic`, `langchain-ollama` to requirements
+- `pyproject.toml` — migrated to `requires-python>=3.13`, added all langchain dev deps
+
+### Fixed
+- ResponseError test updated for new gigachat API signature
+- Dependency resolution: pinned langchain packages to compatible ranges (core<1)
+
+## [0.7.0] - 2026-03-10
+
+### Changed
+- **Project renamed: GigaChain -> SmartChain** — reflects multi-provider nature (not GigaChat-only)
+- Domain: `gigachain` -> `smartchain`
+- Entity classes: `GigaChainConversationEntity` -> `SmartChainConversationEntity`
+- New GitHub repository: `ha-smartchain`
+- HACS name: SmartChain
+- Version bumped to 0.7.0
+
+### Added
+- **AI Task entity** — `SmartChainAITaskEntity` implements `ai_task.AITaskEntity` with `_async_generate_data()` for automation-driven text generation via `ai_task.generate_data` service
+- Structured output support in AI Task (JSON parsing with `task.structure`)
+- Tool calling support in AI Task (reuses conversation entity's LangChain integration)
+
+## [0.6.0] - 2026-03-10
+
+### Added
+- **Assist API for device control** — integration with HA LLM API (`async_provide_llm_data`) allows LLM to call Home Assistant services (turn on/off lights, locks, etc.)
+- `llm_hass_api` option in Options Flow — select HA API for LLM (Assist, custom APIs)
+- HA `llm.Tool` (voluptuous schema) -> LangChain tools conversion via `voluptuous_openapi` + `client.bind_tools()`
+- Tool calling loop with `MAX_TOOL_ITERATIONS = 10`
+- `tool_calls` in `AIMessageChunk` -> HA `ToolInput` in stream deltas
+- `ToolResultContent` <-> LangChain `ToolMessage` conversion in `_chatlog_to_langchain()`
+
+### Changed
+- `_async_handle_message` — uses `async_provide_llm_data` when LLM API configured, manual prompt otherwise
+- Options Flow: `common_config_option_schema` takes `hass` to list available LLM APIs
+
+## [0.5.0] - 2026-03-10
+
+### Added
+- **Streaming responses** — `_attr_supports_streaming = True`, responses streamed via `ChatLog.async_add_delta_content_stream()`
+- Async generator `_async_langchain_stream()` for `AIMessageChunk` -> HA delta dicts
+
+### Changed
+- `client.invoke()` via `async_add_executor_job` replaced with `client.astream()` (async, no executor)
+
+## [0.4.0] - 2026-03-10
+
+### Changed
+- **ChatLog for history** — removed custom `OrderedDict`, uses native HA `ChatLog`
+- **Migration to langchain-gigachat/langchain-openai** — proper package imports
+
+## [0.3.0] - 2026-03-10
+
+### Added
+- **Migration to ConversationEntity** — entity-based conversation agent with `_async_handle_message(user_input, chat_log)`
+
+## [0.2.0] - 2026-03-10
+
+### Fixed
+- Blocking LLM calls, deprecated LangChain API, memory leaks, model defaults
+
+### Removed
+- Anyscale support completely removed
+
+## [0.1.8] - 2024-12-01
+
+### Fixed
+- Compatibility with Home Assistant 2024.12.1+
+
+## [0.1.1] - 2024-03-01
+
+### Added
+- Initial release with GigaChat, YandexGPT support
+- Config Flow and Options Flow
+- Chat history and Jinja2 system prompts