How maksim-sharshov/coredev-fixops works
FixOps monitors labeled containers for structured application errors, then orchestrates code indexing, root-cause analysis, AI patch generation, isolated testing, and repair reporting. After verification, users can apply or roll back repairs; applying can rebuild the workload and commit and push changes. The dashboard presents container health and persisted incident details, with workflow progress streamed live. The README documents the workflow stages and user-visible capabilities; unsampled API, WebSocket, and service implementations limit some lower-level relationship detail.
Main entry points: workflow.py Repair workflow · useAppState.tsx Dashboard state · docker_watcher.py Log watcher · router.py HTTP API
Components 22
Detection and orchestration
- Log watcher
backend/services/docker_watcher.py - Error analysis entry
backend/services/analyze_error.py - Repair workflow
backend/services/workflow.py
Code analysis and repair
- AST indexer
backend/code_intel/indexer.py - Call graph resolver
backend/code_intel/graph.py - Root cause analysis
backend/code_intel/error_analyzer.py - Prompt context builder
backend/code_intel/context_builder.py - DeepSeek client
backend/ai/deepseek.py - Patch and test executor
backend/code_intel/executor.py
Operations and persistence
- HTTP API
backend/api/router.py - WebSocket API
backend/api/ws.py - Container operations
backend/services/container_manager.py - Git integration
backend/services/git_service.py - Incident records
backend/db/psql/models/models.py - Chat history
backend/ai/history_manager.py
Dashboard
- Dashboard pages
frontend/src/App.tsx - Dashboard state
frontend/src/hooks/useAppState.tsx - Incident live view
frontend/src/hooks/useJobSocket.ts - Frontend API client
frontend/src/services/api.ts
Other components
- Monitored application
- Docker workloads
- Developer or operator
Connections 24
- Monitored application → Docker workloads: emits logs
- Docker workloads → Log watcher: streams logs
- Log watcher → Error analysis entry: starts analysis
backend/services/docker_watcher.py - Error analysis entry → Repair workflow: runs workflow
- Repair workflow → AST indexer: indexes project
backend/services/workflow.py - Repair workflow → Call graph resolver: builds call graph
backend/services/workflow.py - Repair workflow → Root cause analysis: locates cause
backend/services/workflow.py - Repair workflow → Prompt context builder: builds prompt
backend/services/workflow.py - Repair workflow → DeepSeek client: requests patch
backend/services/workflow.py - Repair workflow → Patch and test executor: applies and tests
backend/services/workflow.py - Log watcher → WebSocket API: publishes progress
backend/services/docker_watcher.py - Developer or operator → Dashboard pages: views and operates
- Dashboard pages → Dashboard state: provides state
frontend/src/App.tsx - Dashboard state → Frontend API client: loads containers
frontend/src/hooks/useAppState.tsx - Dashboard state → Incident records: loads history via API
- Dashboard state → WebSocket API: receives monitoring events
- Incident live view → WebSocket API: receives job events
- Incident live view → Frontend API client: hydrates incident
frontend/src/hooks/useJobSocket.ts - Frontend API client → HTTP API: requests endpoints
frontend/src/services/api.ts - HTTP API → Container operations: handles container actions
- HTTP API → Incident records: serves incident history
- Container operations → Docker workloads: rebuilds workload
- Container operations → Git integration: syncs and pushes
backend/services/container_manager.py - DeepSeek client → Chat history: stores conversation history
backend/ai/deepseek.py