crca 1.4.0__py3-none-any.whl

prompts/__init__.py

@@ -0,0 +1,10 @@
+"""
+Prompts module for CRCAAgent.
+
+Contains system prompts and instructions for LLM agents using CRCAAgent.
+"""
+
+from .default_crca import DEFAULT_CRCA_SYSTEM_PROMPT
+
+__all__ = ['DEFAULT_CRCA_SYSTEM_PROMPT']
+

prompts/default_crca.py

@@ -0,0 +1,101 @@
+"""
+Default system prompt for CRCAAgent.
+
+This prompt guides LLMs on how to properly use CRCAAgent's causal reasoning
+and counterfactual analysis capabilities, including tool usage.
+"""
+
+DEFAULT_CRCA_SYSTEM_PROMPT = """You are a Causal Reasoning with Counterfactual Analysis (CR-CA) agent assistant.
+
+Your primary role is to help users perform causal reasoning, counterfactual analysis, and understand causal relationships between variables.
+
+## How to Use Tools
+
+**IMPORTANT:** These tools are available via function calling. When you need to use a tool:
+1. The tool will appear in your available functions
+2. Call the function directly with the required parameters
+3. The function will execute and return results
+4. Use those results to continue your analysis
+
+**Do NOT:**
+- Describe what you would do with the tool
+- Ask the user to call the tool
+- Explain the tool without calling it
+
+**DO:**
+- Actually invoke the function when needed
+- Provide all required parameters
+- Use the results returned by the tool
+
+## Available Tools
+
+You have access to two specialized tools that you can call as functions:
+
+### 1. extract_causal_variables
+**When to use:** When the causal graph is empty or you need to identify variables and relationships from a task description.
+
+**How to call the function:**
+This is a function-calling tool. When you need to extract variables, invoke the `extract_causal_variables` function with these parameters:
+  - `required_variables`: List of core variables (e.g., ["price", "demand", "supply"])
+  - `causal_edges`: List of [source, target] pairs showing causal relationships (e.g., [["price", "demand"], ["supply", "price"]])
+  - `reasoning`: Brief explanation of why these variables and relationships are needed
+  - `optional_variables`: (optional) Additional variables that may be useful
+  - `counterfactual_variables`: (optional) Variables to explore in what-if scenarios
+
+**Example:**
+For a task about "predicting product price in 24 months", you might extract:
+- required_variables: ["price", "demand", "supply", "cost", "competition"]
+- causal_edges: [["cost", "price"], ["demand", "price"], ["supply", "price"], ["competition", "price"]]
+- reasoning: "Price is influenced by cost, demand, supply, and competition levels"
+
+**IMPORTANT:** 
+- You MUST call this tool when the causal graph is empty
+- Be thorough - extract ALL relevant variables and relationships
+- You can call this tool multiple times to refine your extraction
+- Think about what factors influence each other in the causal chain
+
+### 2. generate_causal_analysis
+**When to use:** After variables are extracted, to perform detailed causal analysis and counterfactual reasoning.
+
+**How to call the function:**
+This is a function-calling tool. When you need to perform causal analysis, invoke the `generate_causal_analysis` function with these parameters:
+- `causal_analysis`: Detailed analysis of causal relationships and mechanisms
+- `intervention_planning`: Planned interventions to test causal hypotheses
+- `counterfactual_scenarios`: List of scenarios with interventions, expected outcomes, and reasoning
+- `causal_strength_assessment`: Assessment of relationship strengths and confounders
+- `optimal_solution`: Recommended solution based on analysis
+
+## Workflow
+
+1. **Check if variables exist:** If the causal graph is empty, you MUST first call `extract_causal_variables`
+2. **Extract variables:** Use `extract_causal_variables` to identify all relevant variables and causal relationships
+3. **Perform analysis:** Once variables exist, use `generate_causal_analysis` for detailed causal reasoning
+4. **Iterate:** You can refine variable extraction or analysis in multiple rounds
+
+## Best Practices
+
+1. **Be proactive with tool calling:** Don't just describe what you would do - actually call the tools
+2. **Be thorough:** Extract all relevant variables, not just obvious ones
+3. **Think causally:** Consider direct and indirect causal paths
+4. **Validate relationships:** Ensure causal edges make logical sense
+5. **Use reasoning:** Always provide clear reasoning for your variable selections and relationships
+
+## Common Patterns
+
+**Pricing tasks:** Variables like price, demand, supply, cost, competition, market_trends
+**Business tasks:** Variables like revenue, costs, sales, marketing, customer_satisfaction
+**Economic tasks:** Variables like GDP, inflation, employment, interest_rates
+
+## Critical Instructions
+
+- **These are FUNCTION-CALLING tools:** You must actually invoke the functions, not just describe them
+- When you see a task that requires causal analysis, IMMEDIATELY check if variables exist
+- If no variables exist, you MUST call the `extract_causal_variables` function FIRST before any analysis
+- **Do NOT skip function calls** - they are essential for the system to function
+- **Do NOT just describe what you would do** - actually call the functions
+- Always provide complete and valid parameters when calling functions
+- Think step-by-step: identify variables → establish relationships → perform analysis
+- The tools are available in your function calling interface - use them directly
+
+Remember: Your goal is to help users understand causal relationships and make informed decisions through counterfactual analysis."""
+

pyproject.toml

@@ -0,0 +1,151 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "crca"
+version = "1.4.0"
+description = "CR-CA Lite — lightweight causal reasoning agent with advanced analysis suite (CR-CA)"
+readme = "README.md"
+requires-python = ">=3.10"
+license = {text = "Apache-2.0"}
+authors = [
+  { name="Ilum", email="a.leilionas@vk.com" }
+]
+keywords = [
+  "causal-reasoning",
+  "counterfactual-analysis",
+  "structural-causal-models",
+  "llm",
+  "machine-learning",
+  "data-science",
+  "optimization",
+  "governance",
+  "trading",
+  "socioeconomic-dynamics"
+]
+classifiers = [
+  "Development Status :: 4 - Beta",
+  "Intended Audience :: Developers",
+  "Intended Audience :: Science/Research",
+  "License :: OSI Approved :: Apache Software License",
+  "Operating System :: OS Independent",
+  "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",
+  "Topic :: Scientific/Engineering :: Information Analysis",
+  "Topic :: Software Development :: Libraries :: Python Modules",
+]
+dependencies = [
+  "numpy>=1.23",
+  "pandas>=1.5",
+  "scipy>=1.10",
+  "rustworkx>=0.12.0",
+  "networkx>=3.0",
+  "dowhy>=0.8",
+  "optuna>=3.0",
+  "torch>=2.0.0",
+  "scikit-learn>=1.3",
+  "xgboost>=2.0",
+  "lightgbm>=4.0",
+  "numba>=0.58",
+  "yfinance>=0.2.0",
+  "pycoingecko>=3.0",
+  "pytrends>=4.8.0",
+  "praw>=7.0",
+  "shodan>=1.30.0",
+  "geoip2>=4.7.0",
+  "ccxt>=4.0",
+  "websockets>=11.0",
+  "web3>=6.0",
+  "matplotlib>=3.5",
+  "seaborn>=0.11",
+  "pillow>=10.0",
+  "reportlab>=4.0.0",
+  "requests>=2.28",
+  "loguru>=0.7",
+  "python-dotenv>=1.0",
+  "python-dateutil>=2.8",
+  "pyyaml>=6.0",
+  "tqdm>=4.65.0",
+  "rich>=13.0",
+  "redis>=5.0",
+  "pydantic>=2.0",
+  "toml>=0.10.2",
+  "mcp>=0.9.0",
+  "openai>=1.0.0",
+  "swarms",
+  "litellm>=0.1",
+  "psutil>=5.9.0",
+]
+
+[project.optional-dependencies]
+optimization = [
+  "cvxpy>=1.4",  # Requires C++ build tools on Windows (Microsoft Visual C++ 14.0+)
+]
+dev = [
+  "pytest>=7.0",
+  "pytest-cov>=4.0",
+  "black>=23.0",
+  "mypy>=1.0",
+  "ruff>=0.1.0",
+]
+
+[project.urls]
+Homepage = "https://github.com/IlumCI/CR-CA"
+Repository = "https://github.com/IlumCI/CR-CA"
+Documentation = "https://github.com/IlumCI/CR-CA/blob/main/README.md"
+Issues = "https://github.com/IlumCI/CR-CA/issues"
+
+[tool.hatchling.build.targets.wheel]
+# Force hatchling to include everything by using packages configuration
+# Since we have root-level package structure, we explicitly list all packages
+packages = [
+  "utils",
+  "tools",
+  "templates", 
+  "schemas",
+  "branches",
+  "branches.crca_sd",
+  "branches.crca_cg",
+  "prompts",
+]
+# Also explicitly include root files - this ensures they're in the wheel
+# Root files go at wheel root, not in a subdirectory
+shared-data = { "." = ["__init__.py", "CRCA.py"] }
+# Explicitly exclude sensitive files from the package
+exclude = [
+  "**/.env",
+  "**/.env.*",
+  "**/.env.local",
+  "**/*.key",
+  "**/*.pem",
+  "**/secrets.json",
+  "**/credentials.json",
+  "**/config.local.*",
+  "**/*_snapshot_*.json",
+  "**/error.txt",
+  "**/*.log",
+  "**/node_modules/**",
+  "**/venv/**",
+  "**/venv_wsl/**",
+  "**/.git/**",
+  "**/.github/**",
+  "**/__pycache__/**",
+  "**/*.pyc",
+  "**/.pytest_cache/**",
+  "**/dist/**",
+  "**/build/**",
+]
+
+[tool.pytest.ini_options]
+minversion = "7.0"
+addopts = "-q"
+testpaths = ["tests"]
+
+[tool.hatch.build.targets.wheel]
+packages = ["."]
+

requirements.txt

@@ -0,0 +1,76 @@
+# Core Scientific Computing
+numpy>=1.23
+pandas>=1.5
+scipy>=1.10
+
+# Causal Inference & Graph Theory
+rustworkx>=0.12.0
+networkx>=3.0
+dowhy>=0.8
+
+# Optimization
+# Note: cvxpy requires C++ build tools on Windows (Microsoft Visual C++ 14.0+)
+# Install separately if needed: pip install cvxpy
+# Or install build tools: https://visualstudio.microsoft.com/visual-cpp-build-tools/
+# cvxpy>=1.4  # Optional - code handles missing cvxpy gracefully
+optuna>=3.0
+
+# Machine Learning
+torch>=2.0.0
+scikit-learn>=1.3
+xgboost>=2.0
+lightgbm>=4.0
+numba>=0.58
+
+# Data Sources & APIs
+yfinance>=0.2.0
+pycoingecko>=3.0
+pytrends>=4.8.0
+praw>=7.0
+shodan>=1.30.0
+geoip2>=4.7.0
+
+# Trading & Exchanges
+ccxt>=4.0
+websockets>=11.0
+
+# Blockchain
+web3>=6.0
+
+# Visualization
+matplotlib>=3.5
+seaborn>=0.11
+pillow>=10.0
+reportlab>=4.0.0
+
+# Utilities
+requests>=2.28
+loguru>=0.7
+python-dotenv>=1.0
+python-dateutil>=2.8
+pyyaml>=6.0
+tqdm>=4.65.0
+
+# UI
+rich>=13.0
+
+# Caching
+redis>=5.0
+
+# Data Validation & Serialization
+pydantic>=2.0
+toml>=0.10.2
+
+# MCP (Model Context Protocol)
+mcp>=0.9.0
+openai>=1.0.0
+
+# Framework
+swarms
+litellm>=0.1
+
+# System Monitoring
+psutil>=5.9.0
+
+# Optional Dependencies (uncomment if needed)
+# cvxpy>=1.4  # Requires C++ build tools on Windows

schemas/__init__.py

@@ -0,0 +1,43 @@
+"""Schemas module for CR-CA policy engine.
+
+This module provides type definitions and validation for the policy engine.
+"""
+
+from schemas.policy import (
+    DoctrineV1,
+    CompiledPolicy,
+    EpochConfig,
+    MetricSpec,
+    Objective,
+    Invariant,
+    LeverSpec,
+    RiskBudget,
+    ObservationEvent,
+    DecisionEvent,
+    OutcomeEvent,
+    PolicyEvent,
+    LedgerEvent,
+    InterventionSpec,
+    Intervention,
+    ModelState
+)
+
+__all__ = [
+    "DoctrineV1",
+    "CompiledPolicy",
+    "EpochConfig",
+    "MetricSpec",
+    "Objective",
+    "Invariant",
+    "LeverSpec",
+    "RiskBudget",
+    "ObservationEvent",
+    "DecisionEvent",
+    "OutcomeEvent",
+    "PolicyEvent",
+    "LedgerEvent",
+    "InterventionSpec",
+    "Intervention",
+    "ModelState"
+]
+

schemas/mcpSchemas.py

@@ -0,0 +1,51 @@
+"""MCP (Model Context Protocol) schema definitions.
+
+This module provides Pydantic models for MCP connection configurations,
+including single and multiple connection schemas.
+"""
+
+from typing import Any, Dict, List, Optional
+
+from pydantic import BaseModel, Field
+
+
+class MCPConnection(BaseModel):
+    type: Optional[str] = Field(
+        default="mcp",
+        description="The type of connection, defaults to 'mcp'",
+    )
+    url: Optional[str] = Field(
+        default="http://localhost:8000/mcp",
+        description="The URL endpoint for the MCP server",
+    )
+    tool_configurations: Optional[Dict[Any, Any]] = Field(
+        default=None,
+        description="Dictionary containing configuration settings for MCP tools",
+    )
+    authorization_token: Optional[str] = Field(
+        default=None,
+        description="Authentication token for accessing the MCP server",
+    )
+    transport: Optional[str] = Field(
+        default="streamable_http",
+        description="The transport protocol to use for the MCP server",
+    )
+    headers: Optional[Dict[str, str]] = Field(
+        default=None, description="Headers to send to the MCP server"
+    )
+    timeout: Optional[int] = Field(
+        default=10, description="Timeout for the MCP server"
+    )
+
+    class Config:
+        arbitrary_types_allowed = True
+        extra = "allow"
+
+
+class MultipleMCPConnections(BaseModel):
+    connections: List[MCPConnection] = Field(
+        description="List of MCP connections"
+    )
+
+    class Config:
+        arbitrary_types_allowed = True