Changelog
All notable changes to this project will be documented in this file.
[1.1.0] - 2025-12-25
Added
- DataChat MCP Integration: Chat now supports calling external MCP tools alongside SQL queries
- Comprehensive Table Registration: All 44+ project models registered in DataChatRegistry
- Customization Guide: New docs covering admin, developer, and API integration patterns
- automate_api: Full REST API v1 with authentication, pagination, and throttling
- automate_observability: Structured logging, metrics, and OpenTelemetry tracing
- automate_rag: V2 RAG subsystem with Corpus, Document, and Chunk models
- Provider Architecture: New provider registry with entrypoint discovery
- OpenAI Provider: Production-ready OpenAI chat provider implementation
Fixed
- MCP tool sync admin action now correctly unpacks 3-tuple return value
- DataChat prompt template includes MCP tool instructions
- Studio template paths corrected to
admin/automate/studio/ - Admin index.html uses
{% url %}tags instead of hardcoded paths tests/settings.pynow loads.envviapython-dotenv- Database changed from
:memory:to file-based for persistence
Changed
automate_studiomoved to top of INSTALLED_APPS for template precedence- All Studio views now require
staff_member_required automate_datachatre-enabled in INSTALLED_APPS
Documentation
- Added
guides/customization.md- comprehensive customization guide - Updated
features/datachat.mdwith registration examples - Restored full README with architecture diagram and use cases
- Added DataChat table registration docs in apps.py docstring
[1.0.0] - 2025-12-23
Features
- Studio: Visual Automation Wizard, Rule Tester, Execution Explorer
- Core: Reliable Outbox worker, Signal/Webhook triggers
- Governance: SecretRef system, Audit Logging, Policy Engine
- LLM: Gateway with OpenAI/Anthropic support and Jinja2 sandboxing
- Observability: Structured logging and Trace ID propagation
Compatibility
- Django 4.2 / 5.0 / 6.0
- Python 3.10+