How mon1-innovation/maica works
MAICA is a conversational backend with WebSocket and HTTP interfaces, centered on a user-session workflow: prepare context and tools, generate a response, run post-generation triggers and checks, then persist session state. Its documented capabilities also include memory and inspiration features, weather/event/search tools, image understanding, and quality checks. The graph follows relationships shown in excerpts or documented in the README; links to unsampled runtime integrations are omitted or marked unverified.
Main entry points: post_core_pipeliner.py Post-generation pipeline · __init__.py Tool registry · pre_core_pipeliner.py Pre-generation pipeline · database_utils.py Database access
Components 17
Client interfaces
- WebSocket service
maica/maica_ws.py - HTTP service
maica/maica_http.py
Conversation workflow
- Pre-generation pipeline
maica/mfocus/pre_core_pipeliner.py - MFocus agent
maica/mfocus/mfocus_llm.py - MTrigger agent
maica/mtrigger/mtrigger_llm.py - Post-generation pipeline
maica/mtrigger/post_core_pipeliner.py
Assistant capabilities
- Tool registry
maica/mtools/__init__.py - Agent tools
maica/maica_utils/agent_tools.py - Search and events
maica/mtools/serp_get.py - Image understanding
maica/mtools/mvista/img_proc.py - Quality checks
maica/mtools/quality_chk.py - Memory conclusion
maica/mtools/memory_concl.py
State and persistence
- Database access
maica/maica_utils/database_utils.py - Database models
maica/maica_utils/database_models.py - SQL database
Other components
- Chat client
- HTTP client
Connections 14
- Chat client → WebSocket service: connects to
maica/maica_ws.py - HTTP client → HTTP service: sends requests
README.md - Pre-generation pipeline → MFocus agent: runs agent
maica/mfocus/pre_core_pipeliner.py - Pre-generation pipeline → Tool registry: invokes tools
maica/mfocus/pre_core_pipeliner.py - Pre-generation pipeline → Search and events: queries helper
maica/mfocus/pre_core_pipeliner.py - Post-generation pipeline → MTrigger agent: runs trigger
maica/mtrigger/post_core_pipeliner.py - Post-generation pipeline → Quality checks: checks response
maica/mtrigger/post_core_pipeliner.py - Post-generation pipeline → Memory conclusion: concludes memory
maica/mtrigger/post_core_pipeliner.py - Post-generation pipeline → Database access: saves session
- Database access → Database models: uses models
maica/maica_utils/database_utils.py - Database access → SQL database: queries database
maica/maica_utils/database_utils.py - Tool registry → Image understanding: exports capability
maica/mtools/__init__.py - Tool registry → Quality checks: exports capability
maica/mtools/__init__.py - Tool registry → Memory conclusion: exports capability
maica/mtools/__init__.py