How mujeebmohammed-gits/enterprise-pm-agent works
This repository is an early-stage Enterprise PM Agent: its implemented user-to-result path is a client calling the FastAPI service to start or transition a workflow and receive instance or transition data. The workflow engine currently keeps definitions and instances in memory; the storage abstraction exists separately, and the API initializes an unused task store. The dashboard source is a starter screen, not the documented management UI. AI, external-system adapters, mobile, widgets, and the broader security and reporting capabilities are described in the README but are not evidenced by the sampled implementation, so their wiring is omitted.
Main entry points: main.py FastAPI endpoints · engine.py Workflow engine · storage.py Storage adapters · settings.py Runtime settings
Components 8
API service
- FastAPI endpoints
src/api/main.py - Task storage instance
src/api/main.py
Workflow runtime
- Workflow engine
src/core/workflow/engine.py - In-memory workflow state
src/core/workflow/engine.py
State and configuration
- Storage adapters
persistence/storage.py - Runtime settings
config/settings.py
Client
- Starter dashboard
admin-dashboard/src/App.js
Other components
- API user
Connections 7
- API user → FastAPI endpoints: calls REST endpoints
- FastAPI endpoints → Workflow engine: starts or transitions
- FastAPI endpoints → Workflow engine: reads instance state
- Workflow engine → In-memory workflow state: updates workflow state
- FastAPI endpoints → Storage adapters: creates task adapter
- Storage adapters → Task storage instance: instantiates memory store
- FastAPI endpoints → Runtime settings: reads configuration