agent-brain-rag 1.2.0__tar.gz → 2.0.0__tar.gz

{agent_brain_rag-1.2.0 → agent_brain_rag-2.0.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: agent-brain-rag
-Version: 1.2.0
+Version: 2.0.0
 Summary: Agent Brain RAG - Intelligent document indexing and semantic search server that gives AI agents long-term memory
 License: MIT
 Keywords: agent-brain,rag,semantic-search,ai-memory,llm-memory,documentation,indexing,llama-index,chromadb,ai-agent,claude-code,agent-memory
@@ -16,12 +16,19 @@ Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Classifier: Topic :: Software Development :: Documentation
 Classifier: Topic :: Text Processing :: Indexing
+Provides-Extra: graphrag
+Provides-Extra: graphrag-all
+Provides-Extra: graphrag-kuzu
 Requires-Dist: anthropic (>=0.40.0,<0.41.0)
 Requires-Dist: chromadb (>=0.5.0,<0.6.0)
 Requires-Dist: click (>=8.1.0,<9.0.0)
+Requires-Dist: cohere (>=5.0.0,<6.0.0)
 Requires-Dist: fastapi (>=0.115.0,<0.116.0)
+Requires-Dist: google-generativeai (>=0.8.0,<0.9.0)
+Requires-Dist: langextract (>=1.0.0,<2.0.0) ; extra == "graphrag" or extra == "graphrag-all"
 Requires-Dist: llama-index-core (>=0.14.0,<0.15.0)
 Requires-Dist: llama-index-embeddings-openai (>=0.5.0,<0.6.0)
+Requires-Dist: llama-index-graph-stores-kuzu (>=0.9.0,<0.10.0) ; extra == "graphrag-kuzu" or extra == "graphrag-all"
 Requires-Dist: llama-index-llms-openai (>=0.6.12,<0.7.0)
 Requires-Dist: llama-index-readers-file (>=0.5.0,<0.6.0)
 Requires-Dist: llama-index-retrievers-bm25 (>=0.6.0,<0.7.0)
@@ -29,18 +36,25 @@ Requires-Dist: openai (>=1.57.0,<2.0.0)
 Requires-Dist: pydantic (>=2.10.0,<3.0.0)
 Requires-Dist: pydantic-settings (>=2.6.0,<3.0.0)
 Requires-Dist: python-dotenv (>=1.0.0,<2.0.0)
+Requires-Dist: pyyaml (>=6.0.0,<7.0.0)
 Requires-Dist: rank-bm25 (>=0.2.2,<0.3.0)
 Requires-Dist: tiktoken (>=0.8.0,<0.9.0)
 Requires-Dist: tree-sitter-language-pack (>=0.7.3,<0.8.0)
 Requires-Dist: uvicorn[standard] (>=0.32.0,<0.33.0)
-Project-URL: Documentation, https://github.com/SpillwaveSolutions/doc-serve#readme
-Project-URL: Homepage, https://github.com/SpillwaveSolutions/doc-serve
-Project-URL: Repository, https://github.com/SpillwaveSolutions/doc-serve
+Project-URL: Documentation, https://github.com/SpillwaveSolutions/agent-brain/wiki
+Project-URL: Homepage, https://github.com/SpillwaveSolutions/agent-brain
+Project-URL: Repository, https://github.com/SpillwaveSolutions/agent-brain
 Description-Content-Type: text/markdown
 
 # Agent Brain RAG Server
 
-Intelligent document indexing and semantic search REST API service that gives AI agents long-term memory.
+> **Agent Brain** (formerly doc-serve) is an intelligent document indexing and semantic search system designed to give AI agents long-term memory.
+
+AI agents need persistent memory to be truly useful. Agent Brain provides the retrieval infrastructure that enables context-aware, knowledge-grounded AI interactions.
+
+[![PyPI version](https://badge.fury.io/py/agent-brain-rag.svg)](https://pypi.org/project/agent-brain-rag/)
+[![Python 3.10+](https://img.shields.io/badge/python-3.10+-blue.svg)](https://www.python.org/downloads/)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 
 ## Installation
 
@@ -65,28 +79,37 @@ The server will start at `http://127.0.0.1:8000`.
 
 > **Note**: The legacy command `doc-serve` is still available but deprecated. Please use `agent-brain-serve` for new installations.
 
+## Search Capabilities
+
+Agent Brain provides multiple search strategies to match your retrieval needs:
+
+| Search Type | Description | Best For |
+|-------------|-------------|----------|
+| **Semantic Search** | Natural language queries using OpenAI embeddings (`text-embedding-3-large`) | Conceptual questions, finding related content |
+| **Keyword Search (BM25)** | Traditional keyword matching with TF-IDF ranking | Exact matches, technical terms, code identifiers |
+| **Hybrid Search** | Combines vector + BM25 for best of both approaches | General-purpose queries, balanced recall/precision |
+| **GraphRAG** | Knowledge graph-based retrieval for relationship-aware queries | Understanding connections, multi-hop reasoning |
+
 ## Features
 
 - **Document Indexing**: Load and index documents from folders (PDF, Markdown, TXT, DOCX, HTML)
 - **AST-Aware Code Ingestion**: Smart parsing for Python, TypeScript, JavaScript, Java, Go, Rust, C, C++
-- **Hybrid Search**: Combines BM25 keyword matching with semantic similarity
-- **Semantic Search**: Query indexed documents using natural language
+- **Multi-Strategy Retrieval**: Semantic, keyword, hybrid, and graph-based search
 - **OpenAI Embeddings**: Uses `text-embedding-3-large` for high-quality embeddings
+- **Claude Summarization**: AI-powered code summaries for better context
 - **Chroma Vector Store**: Persistent, thread-safe vector database
 - **FastAPI**: Modern, high-performance REST API with OpenAPI documentation
 
-## Quick Start
-
-### Prerequisites
+## Prerequisites
 
 - Python 3.10+
-- Poetry
-- OpenAI API key
+- OpenAI API key (for embeddings)
+- Anthropic API key (for summarization)
 
-### Installation
+## Development Installation
 
 ```bash
-cd doc-serve-server
+cd agent-brain-server
 poetry install
 ```
 
@@ -101,6 +124,7 @@ cp ../.env.example .env
 
 Required environment variables:
 - `OPENAI_API_KEY`: Your OpenAI API key for embeddings
+- `ANTHROPIC_API_KEY`: Your Anthropic API key for summarization
 
 ### Running the Server
 
@@ -112,8 +136,6 @@ poetry run uvicorn agent_brain_server.api.main:app --reload
 poetry run agent-brain-serve
 ```
 
-The server will start at `http://127.0.0.1:8000`.
-
 ### API Documentation
 
 Once running, visit:
@@ -199,6 +221,22 @@ poetry run ruff check agent_brain_server/
 poetry run mypy agent_brain_server/
 ```
 
+## Documentation
+
+- [User Guide](https://github.com/SpillwaveSolutions/agent-brain/wiki/User-Guide) - Getting started and usage
+- [Developer Guide](https://github.com/SpillwaveSolutions/agent-brain/wiki/Developer-Guide) - Contributing and development
+- [API Reference](https://github.com/SpillwaveSolutions/agent-brain/wiki/API-Reference) - Full API documentation
+
+## Release Information
+
+- **Current Version**: See [pyproject.toml](./pyproject.toml)
+- **Release Notes**: [GitHub Releases](https://github.com/SpillwaveSolutions/agent-brain/releases)
+- **Changelog**: [Latest Release](https://github.com/SpillwaveSolutions/agent-brain/releases/latest)
+
+## Related Packages
+
+- [agent-brain-cli](https://pypi.org/project/agent-brain-cli/) - Command-line interface for Agent Brain
+
 ## License
 
 MIT

{agent_brain_rag-1.2.0 → agent_brain_rag-2.0.0}/README.md

@@ -1,6 +1,12 @@
 # Agent Brain RAG Server
 
-Intelligent document indexing and semantic search REST API service that gives AI agents long-term memory.
+> **Agent Brain** (formerly doc-serve) is an intelligent document indexing and semantic search system designed to give AI agents long-term memory.
+
+AI agents need persistent memory to be truly useful. Agent Brain provides the retrieval infrastructure that enables context-aware, knowledge-grounded AI interactions.
+
+[![PyPI version](https://badge.fury.io/py/agent-brain-rag.svg)](https://pypi.org/project/agent-brain-rag/)
+[![Python 3.10+](https://img.shields.io/badge/python-3.10+-blue.svg)](https://www.python.org/downloads/)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 
 ## Installation
 
@@ -25,28 +31,37 @@ The server will start at `http://127.0.0.1:8000`.
 
 > **Note**: The legacy command `doc-serve` is still available but deprecated. Please use `agent-brain-serve` for new installations.
 
+## Search Capabilities
+
+Agent Brain provides multiple search strategies to match your retrieval needs:
+
+| Search Type | Description | Best For |
+|-------------|-------------|----------|
+| **Semantic Search** | Natural language queries using OpenAI embeddings (`text-embedding-3-large`) | Conceptual questions, finding related content |
+| **Keyword Search (BM25)** | Traditional keyword matching with TF-IDF ranking | Exact matches, technical terms, code identifiers |
+| **Hybrid Search** | Combines vector + BM25 for best of both approaches | General-purpose queries, balanced recall/precision |
+| **GraphRAG** | Knowledge graph-based retrieval for relationship-aware queries | Understanding connections, multi-hop reasoning |
+
 ## Features
 
 - **Document Indexing**: Load and index documents from folders (PDF, Markdown, TXT, DOCX, HTML)
 - **AST-Aware Code Ingestion**: Smart parsing for Python, TypeScript, JavaScript, Java, Go, Rust, C, C++
-- **Hybrid Search**: Combines BM25 keyword matching with semantic similarity
-- **Semantic Search**: Query indexed documents using natural language
+- **Multi-Strategy Retrieval**: Semantic, keyword, hybrid, and graph-based search
 - **OpenAI Embeddings**: Uses `text-embedding-3-large` for high-quality embeddings
+- **Claude Summarization**: AI-powered code summaries for better context
 - **Chroma Vector Store**: Persistent, thread-safe vector database
 - **FastAPI**: Modern, high-performance REST API with OpenAPI documentation
 
-## Quick Start
-
-### Prerequisites
+## Prerequisites
 
 - Python 3.10+
-- Poetry
-- OpenAI API key
+- OpenAI API key (for embeddings)
+- Anthropic API key (for summarization)
 
-### Installation
+## Development Installation
 
 ```bash
-cd doc-serve-server
+cd agent-brain-server
 poetry install
 ```
 
@@ -61,6 +76,7 @@ cp ../.env.example .env
 
 Required environment variables:
 - `OPENAI_API_KEY`: Your OpenAI API key for embeddings
+- `ANTHROPIC_API_KEY`: Your Anthropic API key for summarization
 
 ### Running the Server
 
@@ -72,8 +88,6 @@ poetry run uvicorn agent_brain_server.api.main:app --reload
 poetry run agent-brain-serve
 ```
 
-The server will start at `http://127.0.0.1:8000`.
-
 ### API Documentation
 
 Once running, visit:
@@ -159,6 +173,22 @@ poetry run ruff check agent_brain_server/
 poetry run mypy agent_brain_server/
 ```
 
+## Documentation
+
+- [User Guide](https://github.com/SpillwaveSolutions/agent-brain/wiki/User-Guide) - Getting started and usage
+- [Developer Guide](https://github.com/SpillwaveSolutions/agent-brain/wiki/Developer-Guide) - Contributing and development
+- [API Reference](https://github.com/SpillwaveSolutions/agent-brain/wiki/API-Reference) - Full API documentation
+
+## Release Information
+
+- **Current Version**: See [pyproject.toml](./pyproject.toml)
+- **Release Notes**: [GitHub Releases](https://github.com/SpillwaveSolutions/agent-brain/releases)
+- **Changelog**: [Latest Release](https://github.com/SpillwaveSolutions/agent-brain/releases/latest)
+
+## Related Packages
+
+- [agent-brain-cli](https://pypi.org/project/agent-brain-cli/) - Command-line interface for Agent Brain
+
 ## License
 
 MIT

{agent_brain_rag-1.2.0 → agent_brain_rag-2.0.0}/agent_brain_server/__init__.py

@@ -1,3 +1,3 @@
 """Doc-Serve Server - RAG-based document indexing and query service."""
 
-__version__ = "1.2.0"
+__version__ = "2.0.0"

{agent_brain_rag-1.2.0 → agent_brain_rag-2.0.0}/agent_brain_server/api/main.py

@@ -22,6 +22,10 @@ from fastapi.middleware.cors import CORSMiddleware
 
 from agent_brain_server import __version__
 from agent_brain_server.config import settings
+from agent_brain_server.config.provider_config import (
+    load_provider_settings,
+    validate_provider_config,
+)
 from agent_brain_server.indexing.bm25_index import BM25IndexManager
 from agent_brain_server.locking import (
     acquire_lock,
@@ -66,6 +70,30 @@ async def lifespan(app: FastAPI) -> AsyncIterator[None]:
 
     logger.info("Starting Agent Brain RAG server...")
 
+    # Load and validate provider configuration
+    try:
+        provider_settings = load_provider_settings()
+        validation_errors = validate_provider_config(provider_settings)
+
+        if validation_errors:
+            for error in validation_errors:
+                logger.warning(f"Provider config warning: {error}")
+            # Log but don't fail - providers may work if keys are set later
+            # or if using Ollama which doesn't need keys
+
+        # Log active provider configuration
+        logger.info(
+            f"Embedding provider: {provider_settings.embedding.provider} "
+            f"(model: {provider_settings.embedding.model})"
+        )
+        logger.info(
+            f"Summarization provider: {provider_settings.summarization.provider} "
+            f"(model: {provider_settings.summarization.model})"
+        )
+    except Exception as e:
+        logger.error(f"Failed to load provider configuration: {e}")
+        # Continue with defaults - EmbeddingGenerator will handle provider creation
+
     if settings.OPENAI_API_KEY:
         os.environ["OPENAI_API_KEY"] = settings.OPENAI_API_KEY
 
@@ -166,7 +194,7 @@ app = FastAPI(
         "RAG-based document indexing and semantic search API. "
         "Index documents from folders and query them using natural language."
     ),
-    version="1.2.0",
+    version="2.0.0",
     lifespan=lifespan,
     docs_url="/docs",
     redoc_url="/redoc",
@@ -193,7 +221,7 @@ async def root() -> dict[str, str]:
     """Root endpoint redirects to docs."""
     return {
         "name": "Agent Brain RAG API",
-        "version": "1.2.0",
+        "version": "2.0.0",
         "docs": "/docs",
         "health": "/health",
     }

{agent_brain_rag-1.2.0 → agent_brain_rag-2.0.0}/agent_brain_server/api/routers/health.py

@@ -98,4 +98,5 @@ async def indexing_status(request: Request) -> IndexingStatus:
         ),
         indexed_folders=status["indexed_folders"],
         supported_languages=status.get("supported_languages", []),
+        graph_index=status.get("graph_index"),
     )

agent_brain_rag-2.0.0/agent_brain_server/config/provider_config.py

@@ -0,0 +1,308 @@
+"""Provider configuration models and YAML loader.
+
+This module provides Pydantic models for embedding and summarization
+provider configuration, and functions to load configuration from YAML files.
+"""
+
+import logging
+import os
+from functools import lru_cache
+from pathlib import Path
+from typing import Any, Optional
+
+import yaml
+from pydantic import BaseModel, Field, field_validator
+
+from agent_brain_server.providers.base import (
+    EmbeddingProviderType,
+    SummarizationProviderType,
+)
+
+logger = logging.getLogger(__name__)
+
+
+class EmbeddingConfig(BaseModel):
+    """Configuration for embedding provider."""
+
+    provider: EmbeddingProviderType = Field(
+        default=EmbeddingProviderType.OPENAI,
+        description="Embedding provider to use",
+    )
+    model: str = Field(
+        default="text-embedding-3-large",
+        description="Model name for embeddings",
+    )
+    api_key_env: Optional[str] = Field(
+        default="OPENAI_API_KEY",
+        description="Environment variable name containing API key",
+    )
+    base_url: Optional[str] = Field(
+        default=None,
+        description="Custom base URL (for Ollama or compatible APIs)",
+    )
+    params: dict[str, Any] = Field(
+        default_factory=dict,
+        description="Provider-specific parameters",
+    )
+
+    model_config = {"use_enum_values": True}
+
+    @field_validator("provider", mode="before")
+    @classmethod
+    def validate_provider(cls, v: Any) -> EmbeddingProviderType:
+        """Convert string to enum if needed."""
+        if isinstance(v, str):
+            return EmbeddingProviderType(v.lower())
+        if isinstance(v, EmbeddingProviderType):
+            return v
+        return EmbeddingProviderType(v)
+
+    def get_api_key(self) -> Optional[str]:
+        """Resolve API key from environment variable.
+
+        Returns:
+            API key value or None if not found/not needed
+        """
+        if self.provider == EmbeddingProviderType.OLLAMA:
+            return None  # Ollama doesn't need API key
+        if self.api_key_env:
+            return os.getenv(self.api_key_env)
+        return None
+
+    def get_base_url(self) -> Optional[str]:
+        """Get base URL with defaults for specific providers.
+
+        Returns:
+            Base URL for the provider
+        """
+        if self.base_url:
+            return self.base_url
+        if self.provider == EmbeddingProviderType.OLLAMA:
+            return "http://localhost:11434/v1"
+        return None
+
+
+class SummarizationConfig(BaseModel):
+    """Configuration for summarization provider."""
+
+    provider: SummarizationProviderType = Field(
+        default=SummarizationProviderType.ANTHROPIC,
+        description="Summarization provider to use",
+    )
+    model: str = Field(
+        default="claude-haiku-4-5-20251001",
+        description="Model name for summarization",
+    )
+    api_key_env: Optional[str] = Field(
+        default="ANTHROPIC_API_KEY",
+        description="Environment variable name containing API key",
+    )
+    base_url: Optional[str] = Field(
+        default=None,
+        description="Custom base URL (for Grok or Ollama)",
+    )
+    params: dict[str, Any] = Field(
+        default_factory=dict,
+        description="Provider-specific parameters (max_tokens, temperature)",
+    )
+
+    model_config = {"use_enum_values": True}
+
+    @field_validator("provider", mode="before")
+    @classmethod
+    def validate_provider(cls, v: Any) -> SummarizationProviderType:
+        """Convert string to enum if needed."""
+        if isinstance(v, str):
+            return SummarizationProviderType(v.lower())
+        if isinstance(v, SummarizationProviderType):
+            return v
+        return SummarizationProviderType(v)
+
+    def get_api_key(self) -> Optional[str]:
+        """Resolve API key from environment variable.
+
+        Returns:
+            API key value or None if not found/not needed
+        """
+        if self.provider == SummarizationProviderType.OLLAMA:
+            return None  # Ollama doesn't need API key
+        if self.api_key_env:
+            return os.getenv(self.api_key_env)
+        return None
+
+    def get_base_url(self) -> Optional[str]:
+        """Get base URL with defaults for specific providers.
+
+        Returns:
+            Base URL for the provider
+        """
+        if self.base_url:
+            return self.base_url
+        if self.provider == SummarizationProviderType.OLLAMA:
+            return "http://localhost:11434/v1"
+        if self.provider == SummarizationProviderType.GROK:
+            return "https://api.x.ai/v1"
+        return None
+
+
+class ProviderSettings(BaseModel):
+    """Top-level provider configuration."""
+
+    embedding: EmbeddingConfig = Field(
+        default_factory=EmbeddingConfig,
+        description="Embedding provider configuration",
+    )
+    summarization: SummarizationConfig = Field(
+        default_factory=SummarizationConfig,
+        description="Summarization provider configuration",
+    )
+
+
+def _find_config_file() -> Optional[Path]:
+    """Find the configuration file in standard locations.
+
+    Search order:
+    1. DOC_SERVE_CONFIG environment variable
+    2. Current directory config.yaml
+    3. State directory config.yaml (if DOC_SERVE_STATE_DIR set)
+    4. Project root config.yaml
+
+    Returns:
+        Path to config file or None if not found
+    """
+    # 1. Environment variable override
+    env_config = os.getenv("DOC_SERVE_CONFIG")
+    if env_config:
+        path = Path(env_config)
+        if path.exists():
+            return path
+        logger.warning(f"DOC_SERVE_CONFIG points to non-existent file: {env_config}")
+
+    # 2. Current directory
+    cwd_config = Path.cwd() / "config.yaml"
+    if cwd_config.exists():
+        return cwd_config
+
+    # 3. State directory
+    state_dir = os.getenv("DOC_SERVE_STATE_DIR")
+    if state_dir:
+        state_config = Path(state_dir) / "config.yaml"
+        if state_config.exists():
+            return state_config
+
+    # 4. .claude/doc-serve directory (project root pattern)
+    claude_dir = Path.cwd() / ".claude" / "doc-serve"
+    if claude_dir.exists():
+        claude_config = claude_dir / "config.yaml"
+        if claude_config.exists():
+            return claude_config
+
+    return None
+
+
+def _load_yaml_config(path: Path) -> dict[str, Any]:
+    """Load YAML configuration from file.
+
+    Args:
+        path: Path to YAML config file
+
+    Returns:
+        Configuration dictionary
+
+    Raises:
+        ConfigurationError: If YAML parsing fails
+    """
+    from agent_brain_server.providers.exceptions import ConfigurationError
+
+    try:
+        with open(path) as f:
+            config = yaml.safe_load(f)
+            return config if config else {}
+    except yaml.YAMLError as e:
+        raise ConfigurationError(
+            f"Failed to parse config file {path}: {e}",
+            "config",
+        ) from e
+    except OSError as e:
+        raise ConfigurationError(
+            f"Failed to read config file {path}: {e}",
+            "config",
+        ) from e
+
+
+@lru_cache
+def load_provider_settings() -> ProviderSettings:
+    """Load provider settings from YAML config or defaults.
+
+    This function:
+    1. Searches for config.yaml in standard locations
+    2. Parses YAML and validates against Pydantic models
+    3. Falls back to defaults (OpenAI embeddings + Anthropic summarization)
+
+    Returns:
+        Validated ProviderSettings instance
+    """
+    config_path = _find_config_file()
+
+    if config_path:
+        logger.info(f"Loading provider config from {config_path}")
+        raw_config = _load_yaml_config(config_path)
+        settings = ProviderSettings(**raw_config)
+    else:
+        logger.info("No config file found, using default providers")
+        settings = ProviderSettings()
+
+    # Log active configuration
+    logger.info(
+        f"Active embedding provider: {settings.embedding.provider} "
+        f"(model: {settings.embedding.model})"
+    )
+    logger.info(
+        f"Active summarization provider: {settings.summarization.provider} "
+        f"(model: {settings.summarization.model})"
+    )
+
+    return settings
+
+
+def clear_settings_cache() -> None:
+    """Clear the cached provider settings (for testing)."""
+    load_provider_settings.cache_clear()
+
+
+def validate_provider_config(settings: ProviderSettings) -> list[str]:
+    """Validate provider configuration and return list of errors.
+
+    Checks:
+    - API keys are available for providers that need them
+    - Models are known for the selected provider
+
+    Args:
+        settings: Provider settings to validate
+
+    Returns:
+        List of validation error messages (empty if valid)
+    """
+    errors: list[str] = []
+
+    # Validate embedding provider
+    if settings.embedding.provider != EmbeddingProviderType.OLLAMA:
+        api_key = settings.embedding.get_api_key()
+        if not api_key:
+            env_var = settings.embedding.api_key_env or "OPENAI_API_KEY"
+            errors.append(
+                f"Missing API key for {settings.embedding.provider} embeddings. "
+                f"Set {env_var} environment variable."
+            )
+
+    # Validate summarization provider
+    if settings.summarization.provider != SummarizationProviderType.OLLAMA:
+        api_key = settings.summarization.get_api_key()
+        if not api_key:
+            env_var = settings.summarization.api_key_env or "ANTHROPIC_API_KEY"
+            errors.append(
+                f"Missing API key for {settings.summarization.provider} summarization. "
+                f"Set {env_var} environment variable."
+            )
+
+    return errors

{agent_brain_rag-1.2.0 → agent_brain_rag-2.0.0}/agent_brain_server/config/settings.py

@@ -26,7 +26,7 @@ class Settings(BaseSettings):
 
     # Anthropic Configuration
     ANTHROPIC_API_KEY: str = ""
-    CLAUDE_MODEL: str = "claude-3-5-haiku-20241022"  # Claude 3.5 Haiku (latest)
+    CLAUDE_MODEL: str = "claude-haiku-4-5-20251001"  # Claude 4.5 Haiku (latest)
 
     # Chroma Configuration
     CHROMA_PERSIST_DIR: str = "./chroma_db"
@@ -51,6 +51,17 @@ class Settings(BaseSettings):
     DOC_SERVE_STATE_DIR: Optional[str] = None  # Override state directory
     DOC_SERVE_MODE: str = "project"  # "project" or "shared"
 
+    # GraphRAG Configuration (Feature 113)
+    ENABLE_GRAPH_INDEX: bool = False  # Master switch for graph indexing
+    GRAPH_STORE_TYPE: str = "simple"  # "simple" (in-memory) or "kuzu" (persistent)
+    GRAPH_INDEX_PATH: str = "./graph_index"  # Path for graph persistence
+    GRAPH_EXTRACTION_MODEL: str = "claude-haiku-4-5"  # Model for entity extraction
+    GRAPH_MAX_TRIPLETS_PER_CHUNK: int = 10  # Max triplets per document chunk
+    GRAPH_USE_CODE_METADATA: bool = True  # Use AST metadata for code entities
+    GRAPH_USE_LLM_EXTRACTION: bool = True  # Use LLM for additional extraction
+    GRAPH_TRAVERSAL_DEPTH: int = 2  # Depth for graph traversal in queries
+    GRAPH_RRF_K: int = 60  # Reciprocal Rank Fusion constant for multi-retrieval
+
     model_config = SettingsConfigDict(
         env_file=[
             ".env",  # Current directory

{agent_brain_rag-1.2.0 → agent_brain_rag-2.0.0}/agent_brain_server/indexing/__init__.py

@@ -7,6 +7,18 @@ from agent_brain_server.indexing.embedding import (
     EmbeddingGenerator,
     get_embedding_generator,
 )
+from agent_brain_server.indexing.graph_extractors import (
+    CodeMetadataExtractor,
+    LLMEntityExtractor,
+    get_code_extractor,
+    get_llm_extractor,
+    reset_extractors,
+)
+from agent_brain_server.indexing.graph_index import (
+    GraphIndexManager,
+    get_graph_index_manager,
+    reset_graph_index_manager,
+)
 
 __all__ = [
     "DocumentLoader",
@@ -16,4 +28,13 @@ __all__ = [
     "get_embedding_generator",
     "BM25IndexManager",
     "get_bm25_manager",
+    # Graph indexing (Feature 113)
+    "LLMEntityExtractor",
+    "CodeMetadataExtractor",
+    "get_llm_extractor",
+    "get_code_extractor",
+    "reset_extractors",
+    "GraphIndexManager",
+    "get_graph_index_manager",
+    "reset_graph_index_manager",
 ]