Hillstar Orchestrator

Security and reproducibility-first workflow orchestrator for research environments.

Hillstar is a research-grade workflow orchestration system designed for multi-agent AI pipelines and classification tasks. It provides a modular architecture with support for multiple LLM providers, governance policies, and comprehensive execution tracing.

API Reference

• Configuration & Provider Registry
  • config.config
  • config.model_selector
  • config.provider_registry
  • config.setup_wizard
• Workflow Execution
  • execution.runner
  • execution.graph
  • execution.node_executor
  • execution.trace
  • execution.checkpoint
  • execution.observability
  • execution.config_validator
  • execution.cost_manager
  • execution.model_selector
• Compliance & Governance
  • GovernanceEnforcer
  • HookManager
  • GovernancePolicy
  • verify_hillstar_compliance()
  • governance.enforcer
  • governance.compliance
  • governance.policy
  • governance.hooks
  • governance.project_init
• LLM Model Providers
  • AnthropicOllamaAPIModel
  • AnthropicModel
  • DevstralLocalModel
  • MistralAPIModel
  • MCPModel
  • AnthropicMCPModel
  • OpenAIMCPModel
  • MistralMCPModel
  • OllamaMCPModel
  • models.mcp_model
  • models.anthropic_model
  • models.anthropic_mcp_model
  • models.anthropic_ollama_api_model
  • models.mistral_api_model
  • models.mistral_mcp_model
  • models.openai_mcp_model
  • models.ollama_mcp_model
  • models.devstral_local_model
• Utilities
  • HillstarException
  • ConfigurationError
  • BudgetExceededError
  • ModelSelectionError
  • redact()
  • contains_credentials()
  • CredentialRedactor
  • utils.credential_redactor
  • utils.exceptions
  • utils.json_output_viewer
  • utils.report
• Workflow Discovery & Validation
  • WorkflowValidator
  • ModelPresets
  • AutoDiscover
  • WorkflowDiscovery
  • workflows.discovery
  • workflows.auto_discover
  • workflows.validator
  • workflows.model_presets
• MCP Servers
  • MCP Server: Base Class for All Providers
  • MCP Server: Anthropic Claude Models
  • MCP Server: Fallback to Ollama Cloud Models via claude-ollama CLI
  • MCP Server: Devstral Local (llama.cpp HTTP Server)
  • MCP Server: File Operations (write_file, update_file, create_directory)
  • MCP Server: Google AI Studio (Gemini Models)
  • MCP Server: Mistral AI Models
  • MCP Server: Ollama Local Models
  • MCP Server: OpenAI GPT Models with Dual Authentication
  • Secure logging module for MCP servers.

Top-Level

• Command-Line Interface
• Package Root

Indices and tables

• Index

• Module Index

• Search Page