How pranavubarhande/vivexy works
Vivexy is an AI operating system for professional teams, delivered through web and desktop clients backed by a Flask service and shared agent runtime. A user configures an agent, submits a chat or document task, and receives streamed answers, citations or generated artifacts; the platform also supports domain tools, RAG, scheduled runs, and team controls. The sampled sources show scheduled-task API dispatch and desktop MCP handling directly. Other major capabilities are retained from the README, but their internal call paths were not sampled, so those links are marked unverified.
Main entry points: chat.py Chat orchestration · base.py Application database · scheduled_tasks.py Schedule management · app.py Flask API
Components 25
User clients
- Web workspace
packages/core/src/components/chat/agentic-chat.tsx - Desktop app
packages/desktop/src/ipc-handlers.ts - Authentication
packages/core/src/services/auth.service.ts - Desktop MCP manager
packages/desktop/src/tools/mcp-manager.ts
Service API
- Flask API
apps/agents-service/app.py - Chat and agent routes
packages/agents-core/src/agents_core/routes/chat.py - Document analysis routes
packages/agents-core/src/agents_core/routes/document_analyses.py - Document and RAG routes
packages/agents-core/src/agents_core/routes/rag.py - Scheduled-task routes
packages/agents-core/src/agents_core/routes/scheduled_tasks.py
Agent runtime
- Chat orchestration
packages/agents-core/src/agents_core/business_logic/chat.py - Agent configuration
packages/agents-core/src/agents_core/agents/config.py - Streaming responses
packages/agents-core/src/agents_core/business_logic/streaming.py - Job queues
packages/agents-core/src/agents_core/worker/dispatcher.py - Chat workers
packages/agents-core/src/agents_core/worker/consumer.py
Knowledge and tools
- Retrieval service
packages/agents-core/src/agents_core/services/rag_service.py - Document processing
packages/agents-core/src/agents_core/services/file_processor.py - Domain tools
packages/agents-core/src/agents_core/tools - Skills and MCP
packages/agents-core/src/agents_core/agents/mcp_client.py
Team operations
- Schedule management
packages/agents-core/src/agents_core/business_logic/scheduled_tasks.py - Task scheduler
packages/agents-core/src/agents_core/worker/scheduler.py - Scheduled task runner
packages/agents-core/src/agents_core/worker/scheduled_task_runner.py - Teams and policy
packages/agents-core/src/agents_core/business_logic/teams.py - Billing and usage
packages/agents-core/src/agents_core/business_logic/billing.py - Application database
packages/agents-core/src/agents_core/models/base.py
Other components
- Professional user
Connections 27
- Professional user → Web workspace: uses
- Professional user → Desktop app: uses
- Web workspace → Flask API: requests
- Authentication → Flask API: authenticates
packages/core/src/services/auth.service.ts - Flask API → Chat and agent routes: routes requests
- Chat and agent routes → Chat orchestration: delegates
- Chat orchestration → Agent configuration: configures agents
- Chat orchestration → Job queues: dispatches jobs
- Chat workers → Chat orchestration: runs jobs
- Chat orchestration → Streaming responses: streams results
- Chat orchestration → Retrieval service: retrieves context
- Chat orchestration → Domain tools: invokes tools
- Chat orchestration → Skills and MCP: loads skills
- Document and RAG routes → Document processing: processes uploads
- Document and RAG routes → Retrieval service: retrieves documents
- Document analysis routes → Document processing: analyzes documents
- Scheduled-task routes → Schedule management: calls task logic
packages/agents-core/src/agents_core/routes/scheduled_tasks.py - Schedule management → Job queues: dispatches runs
packages/agents-core/src/agents_core/business_logic/scheduled_tasks.py - Task scheduler → Scheduled task runner: triggers runs
- Scheduled task runner → Chat orchestration: runs prompt
- Schedule management → Application database: stores schedules
packages/agents-core/src/agents_core/business_logic/scheduled_tasks.py - Chat orchestration → Application database: stores conversations
- Document processing → Application database: stores document data
- Teams and policy → Application database: stores team state
- Billing and usage → Application database: records usage
- Desktop app → Desktop MCP manager: starts MCP
packages/desktop/src/ipc-handlers.ts - Desktop MCP manager → Skills and MCP: proxies tool calls