Clean Up of Dev Environment (#15)

* removed mentions of neptune
* simplified dependency management
* bumped dev package versions

Author: cemde

.github/workflows/test.yml

@@ -67,7 +67,7 @@ jobs:
       - name: Install all dependencies
         run: |
           pip install uv
-          uv sync --group dev --group typing --group examples
+          uv sync --all-extras --group dev
       - name: Run all tests
         run: |
           uv run pytest -v

.github/workflows/ty-check.yml

@@ -27,11 +27,8 @@ jobs:
       - name: Install uv
         run: python -m pip install uv
 
-      - name: Install dependencies with typing group
-        run: uv sync --group typing
-
-      - name: Install ty type checker
-        run: uv pip install "ty>=0.0.1a20"
+      - name: Install dependencies with all optional packages for type checking
+        run: uv sync --all-extras --group dev
 
       - name: Run ty checks
         run: |

AGENTS.md

@@ -60,17 +60,17 @@ uv run pytest -m interface -v
 Three types of dependencies:
 
 - **Core** (`[project.dependencies]`): Required, installed with `pip install maseval`. Keep minimal!
-- **Optional** (`[project.optional-dependencies]`): Published for end users. Framework integrations like `smolagents`, `langgraph`.
+- **Optional** (`[project.optional-dependencies]`): Published for end users. Framework integrations like `smolagents`, `langgraph`, plus convenience bundles like `all` and `examples`. Uses self-references (e.g., `maseval[all]`) to avoid duplication - this is a standard Python packaging feature.
 - **Dev Groups** (`[dependency-groups]`): NOT published. Only for contributors. Tools like `pytest`, `ruff`, `mkdocs`.
 
 ```bash
 # Add core dependency (use sparingly!)
 uv add <package-name>
 
-# Add optional dependency for end users
+# Add optional dependency for end users (e.g., framework integrations)
 uv add --optional <extra-name> <package-name>
 
-# Add development dependency (not published)
+# Add development dependency (not published - dev tools only)
 uv add --group dev <package-name>
 
 # Remove any dependency
@@ -79,6 +79,17 @@ uv remove <package-name>
 
 **Important:** `uv add` automatically updates both `pyproject.toml` and `uv.lock`. Never edit lockfile manually.
 
+**Understanding `uv sync` options:**
+
+- `uv sync`: Installs only core dependencies
+- `uv sync --extra <name>`: Adds specific optional dependency (e.g., `--extra smolagents`)
+- `uv sync --all-extras`: Installs ALL optional dependencies (includes `smolagents`, `langgraph`, etc.)
+- `uv sync --group <name>`: Adds specific dev group (e.g., `--group dev`)
+- `uv sync --all-groups`: Installs ALL dev groups (`dev`, `docs`)
+- `uv sync --all-extras --all-groups`: Full contributor setup with everything
+
+**Note:** Type checking requires `--all-extras` to install all framework packages needed for checking types across all interfaces.
+
 ## Architecture Rules
 
 **Core vs Interface Separation (CRITICAL)**

maseval/core/utils/system_info.py

@@ -164,7 +164,7 @@ def get_environment_variables(include_patterns: Optional[list[str]] = None) -> D
         include_patterns: Optional list of substring patterns to match against environment variable names (case-insensitive).
             If None, defaults to common AI/ML framework variables including:
             - AI Services: OPENAI, ANTHROPIC, GOOGLE, GEMINI, HUGGINGFACE, HF_
-            - Experiment Tracking: WANDB, NEPTUNE, LANGFUSE
+            - Experiment Tracking: WANDB, LANGFUSE
             - ML Frameworks: CUDA, PYTORCH, TF_ (TensorFlow)
             - Python Environment: PATH, PYTHONPATH, VIRTUAL_ENV, CONDA
 
@@ -183,7 +183,6 @@ def get_environment_variables(include_patterns: Optional[list[str]] = None) -> D
             "HUGGINGFACE",
             "HF_",
             "WANDB",
-            "NEPTUNE",
             "LANGFUSE",
             "CUDA",
             "PYTORCH",

maseval/interface/__init__.py

@@ -6,7 +6,7 @@
 Organization:
 - inference/: Model inference adapters (OpenAI, Google, HuggingFace, etc.)
 - agents/: Agent framework adapters (smolagents, langgraph, etc.)
-- logging/: Logging platform adapters (wandb, neptune, langfuse, etc.)
+- logging/: Logging platform adapters (wandb, langfuse, etc.)
 
 Canonical rules:
 - Keep adapters thin: translate between MASEval internal abstractions and the external API.

maseval/interface/logging/__init__.py

@@ -5,4 +5,4 @@
 
 __all__ = []
 
-# TODO: Add wandb, neptune, langfuse adapters
+# TODO: Add wandb, langfuse adapters

pyproject.toml

@@ -8,91 +8,54 @@ version = "0.2.0"
 description = "A library for multi-agent systems evaluation and benchmarking."
 authors = [{ name = "Cornelius Emde", email = "cornelius@parameterlab.de" }]
 readme = "README.md"
+license = { file = "LICENSE" }
 requires-python = ">=3.10"
 classifiers = [
     "Development Status :: 3 - Alpha",
     "Intended Audience :: Science/Research",
+    "License :: OSI Approved :: MIT License",
     "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
     "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
     "Topic :: Scientific/Engineering :: Artificial Intelligence",
 ]
 dependencies = ["gitpython>=3.1.0", "tqdm>=4.66.0", "rich>=14.1.0"]
 
-# Enable optional dependencies
+# Enable optional dependencies for end users
 [project.optional-dependencies]
 # Agent frameworks
-smolagents = ["smolagents>=1.21"]
+smolagents = ["smolagents>=1.21.3"]
 langgraph = ["langgraph>=0.6.0"]
 llamaindex = ["llama-index-core>=0.12.0"]
 
-# Inference Enginges
+# Inference engines
 openai = ["openai>=1.107.2"]
 google-genai = ["google-genai>=1.37.0"]
 transformers = ["transformers>=4.37.0"]
-litellm = ["litellm>=0.2.0"]
+litellm = ["litellm>=1.0.0"]
 
 # Logging
 wandb = ["wandb>=0.15.0"]
-neptune = ["neptune>=1.14.0"]
 langfuse = ["langfuse>=3.3.4"]
 
-# Combinations
+# Dependencies for running examples (only what's actually used)
 examples = [
-    "datasets>=4.0.0",
-    "google-genai>=1.37.0",
+    "maseval[smolagents,langgraph,llamaindex,openai,google-genai,litellm,langfuse]",
+    # Additional integrations used in examples
     "langchain>=0.3.27",
-    "langchain-litellm>=0.1.0",
     "langchain-google-genai>=2.1.12",
-    "langgraph>=0.6.0",
-    "langfuse>=3.3.4",
-    "neptune>=1.14.0",
-    "openai>=1.107.2",
-    "smolagents>=1.21.3",
-    "llama-index-core>=0.12.0",
-    "llama-index-llms-litellm>=0.3.0",
-    "litellm>=1.0.0",
-    "addict>=2.4.0",
-    "deepdiff>=8.6.1",
-    "docstring-parser>=0.17.0",
-    "loguru>=0.7.3",
-    "toml>=0.10.2",
     "typing-extensions>=4.0.0",
-    "restrictedpython>=8.1",
     "mcp>=1.22.0",
     # Jupyter notebook support
     "jupyter>=1.0.0",
     "ipykernel>=6.0.0",
     "ipywidgets>=8.0.0",
 ]
-all = [
-    # Agent frameworks
-    "smolagents>=1.21",
-    "langgraph>=0.6.0",
-    "llama-index-core>=0.12.0",
-    # Inference engines
-    "openai>=1.107.2",
-    "google-genai>=1.37.0",
-    "transformers>=4.37.0",
-    "litellm>=0.2.0",
-    # Logging
-    "wandb>=0.15.0",
-    "neptune>=1.14.0",
-    "langfuse>=3.3.4",
-    # Examples dependencies
-    "datasets>=4.0.0",
-    "langchain>=0.3.27",
-    "langchain-litellm>=0.1.0",
-    "langchain-google-genai>=2.1.12",
-    "addict>=2.4.0",
-    "deepdiff>=8.6.1",
-    "docstring-parser>=0.17.0",
-    "loguru>=0.7.3",
-    "toml>=0.10.2",
-    "typing-extensions>=4.0.0",
-    "jupyter>=1.0.0",
-    "ipykernel>=6.0.0",
-    "ipywidgets>=8.0.0",
-]
+
+# Complete installation with absolutely everything (uses self-reference for DRY)
+all = ["maseval[examples,transformers,wandb]"]
 
 [project.urls]
 "Homepage" = "https://github.com/parameterlab/MASEval"
@@ -101,64 +64,19 @@ all = [
 "Changelog" = "https://github.com/parameterlab/MASEval/blob/main/CHANGELOG.md"
 
 [dependency-groups]
-# Dependencies needed for development (linting, testing)
-dev = ["pytest>=8.4.2", "pytest-cov>=7.0.0", "ruff>=0.13.0", "ty>=0.0.1a20"]
-
+# Development tools (linting, testing) - for contributors only
+dev = ["pytest>=9.0.0", "pytest-cov>=7.0.0", "ruff>=0.14.0", "ty>=0.0.5"]
 
-# Dependencies needed to build the documentation
+# Documentation building - for contributors only
 docs = [
-    "mkdocs>=1.5",
-    "mkdocs-material>=9.0.0",
-    "mkdocstrings[python]>=0.17.0",
-    "pymdown-extensions>=9.0.0",
+    "mkdocs>=1.6",
+    "mkdocs-material>=9.7.0",
+    "mkdocstrings[python]>=1.0.0",
+    "pymdown-extensions>=10.0.0",
     "mkdocs-jupyter>=0.24.0",
     "mkdocs-git-revision-date-localized-plugin>=1.5.0",
 ]
 
-# Dependencies needed for static type checking
-typing = [
-    # All optional dependencies for complete type checking
-    "smolagents>=1.21",
-    "langgraph>=0.6.0",
-    "llama-index-core>=0.12.0",
-    "openai>=1.107.2",
-    "google-genai>=1.37.0",
-    "transformers>=4.37.0",
-    "litellm>=0.2.0",
-    "wandb>=0.15.0",
-    "neptune>=1.14.0",
-    "langfuse>=3.3.4",
-]
-
-# Dependencies needed to run the example scripts (for contributors)
-examples = [
-    "datasets>=4.0.0",
-    "google-genai>=1.37.0",
-    "langchain>=0.3.27",
-    "langfuse>=3.3.4",
-    "neptune>=1.14.0",
-    "openai>=1.107.2",
-    "smolagents>=1.21.3",
-    "langchain-google-genai>=2.1.12",
-    "langgraph>=0.6.0",
-    "llama-index-core>=0.12.0",
-    "llama-index-llms-litellm>=0.3.0",
-    "litellm>=1.0.0",
-    "addict>=2.4.0",
-    "deepdiff>=8.6.1",
-    "docstring-parser>=0.17.0",
-    "loguru>=0.7.3",
-    "toml>=0.10.2",
-    "typing-extensions>=4.0.0",
-    "restrictedpython>=8.1",
-    "mcp>=1.22.0",
-    # Jupyter notebook support
-    "jupyter>=1.0.0",
-    "ipykernel>=6.0.0",
-    "ipywidgets>=8.0.0",
-    "openinference-instrumentation-smolagents>=0.1.20",
-]
-
 [tool.setuptools]
 packages = ["maseval"]