Feature: Supervisor Workflow + DI #88
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…pervisor-workflow
Task 10 cleanup: - Remove event_coordinator import/usage from agents/__init__.py - Delete tests/test_workflow_executors.py (obsolete coordinator tests) - Delete tests/test_workflow_no_summarization.py (obsolete coordinator tests) - Update budget_analyst.py to remove SpecialistOutput reference in comment - Update supervisor.py to remove EventPlanningCoordinator reference in docstring All remaining tests pass (45 passed, 5 skipped). 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
- Remove unused global_tools parameters from agent factory functions - Update imports and dependencies throughout codebase - Fix test imports and assertions to match new supervisor pattern - Clean up docstrings to remove obsolete parameter documentation This completes the removal of the old coordinator pattern in favor of the new supervisor-based workflow architecture. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
lavaman131
changed the title
Contributor
There was a problem hiding this comment.
Pull Request Overview
This PR implements a major architectural refactoring from a coordinator-centric pattern to a supervisor pattern with dependency injection:
- Replaces
EventPlanningCoordinatorwithSupervisorOrchestratorExecutorwhere routing decisions are made by an LLM supervisor agent via structured output - Introduces dependency injection using
dependency-injectorlibrary to manage client and global tools lifecycle - Refactors all agent factory functions to use
@injectdecorator for automatic dependency injection - Updates
SpecialistOutputmodel toSupervisorDecisionmodel for routing decisions - Removes deleted files including old coordinator code and obsolete test files
Reviewed Changes
Copilot reviewed 38 out of 38 changed files in this pull request and generated 2 comments.
Show a summary per file
| File | Description |
|---|---|
| src/spec_to_agents/workflow/supervisor.py | New supervisor orchestrator and workflow builder for dynamic routing |
| src/spec_to_agents/container.py | New DI container managing client and global tools |
| src/spec_to_agents/models/messages.py | Renamed SpecialistOutput to SupervisorDecision |
| src/spec_to_agents/agents/*.py | Refactored agent factories to use @inject for DI |
| src/spec_to_agents/workflow/core.py | Updated to use supervisor pattern with DI |
| src/spec_to_agents/console.py | Updated to initialize DI container |
| tests/*.py | Updated tests for new architecture with type hints |
| pyproject.toml | Added dependency-injector and langchain-core dependencies |
thegovind
reviewed
Nov 2, 2025
Collaborator
thegovind
left a comment
thegovind
left a comment
There was a problem hiding this comment.
I think we should remove langchain dependency here - @lavaman131 @copilot
Contributor
|
@thegovind I've opened a new pull request, #89, to work on those changes. Once the pull request is ready, I'll request review from you. |
thegovind
reviewed
Nov 2, 2025
markwallace-microsoft
approved these changes
Nov 2, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Short Description
Implements new supervisor workflow that simplifies event planning system. For details on the design refer to the spec here.
Modifies agent creation to use dependency injection and local tool creation in agent factory functions.
Long Description
This pull request refactors agent creation to use dependency injection and async context management, improving lifecycle handling, testability, and code clarity. The changes remove global state and manual cleanup for the MCP sequential-thinking tool, update agent constructors to use injected dependencies, and modernize the project’s architecture for better composability and maintainability.
Dependency Injection & Agent Construction:
budget_analyst,catering_coordinator,logistics_manager,venue_specialist) to use dependency injection viadependency-injector. Agent constructors now receive their dependencies automatically, removing the need to pass clients or tools manually. [1] [2] [3] [4]clientandglobal_toolsvia DI, and include the MCP sequential-thinking tool only if present inglobal_tools. This enables explicit dependency injection and easier testing/mocking. [1] [2] [3] [4]Removal of Event Coordinator Agent Export:
event_coordinatoragent is no longer exported fromagents/__init__.pyand its module has been removed, reflecting a shift to specialist agents only. Theexport_agents()function now returns only specialist agents and documents DI requirements. [1] [2] [3]Async Context Management for MCP Tool:
Dependency Updates:
dependency-injectorandlangchain-coretopyproject.tomldependencies to support the new DI pattern and future extensibility.These changes together modernize agent construction, remove legacy patterns, and lay the foundation for robust, scalable multi-agent workflows.