How dreamsartificial648-sketch/adam-architecture-v2 works
ADAM is a desktop AI creation and training studio: users describe or configure work, the planner turns requests into plans, and jobs execute registered tools whose outputs become models, generated media, or datasets. The tree also indicates dedicated dataset/Oasis workflows, model and generation features, experiment tracking, and authenticated remote control. Wiring is clearest for application construction and job execution; several feature modules were not sampled, so their internal call paths are intentionally not inferred.
Main entry points: main_window.py Studio UI · app.py Application setup · remote_access.py Remote API · executor.py Tool executor
Components 22
Application
- Studio UI
main_window.py - Application setup
app.py
Planning and Jobs
- Request planner
planner.py - Job manager
job_manager.py - Tool executor
executor.py - Tool registry
registry.py
AI Workflows
- Training workflows
studio.py - Image generation
generations.py - Live previews
generation_previews.py - Oasis gameplay
oasis_player.py - Frame reviewer
cnn_reviewer.py - Web search
web_search.py
Assets and Remote
- Dataset workflows
dataset_lab.py - Dataset collection
real_dataset_collector.py - Asset registry
assets.py - Experiment tracking
experiment_tracker.py - Remote API
remote_access.py - Remote commands
remote_dispatcher.py - Remote dashboard
remote_dashboard.py
Other components
- Studio user
- Remote client
- Registered tools
Connections 21
- Studio user → Studio UI: uses
- Application setup → Studio UI: constructs
- Application setup → Request planner: constructs
- Application setup → Tool registry: constructs
- Application setup → Tool executor: constructs
- Application setup → Job manager: constructs
- Studio UI → Request planner: plans requests
- Studio UI → Job manager: controls jobs
- Request planner → Tool registry: looks up tools
- Job manager → Tool executor: runs steps
- Tool executor → Tool registry: resolves tools
- Tool executor → Registered tools: dispatches work
- Studio UI → Image generation: opens workflow
- Studio UI → Dataset workflows: opens workflow
- Studio UI → Remote API: configures access
- Remote client → Remote dashboard: uses
- Remote dashboard → Remote API: requests API
- Remote API → Remote commands: dispatches commands
- Remote API → Job manager: reads job state
- Remote API → Request planner: plans requests
- Frame reviewer → Asset registry: stores checkpoint