# arjitbhatnagar3-hue/meta-agent-learning architecture

> Architecture diagram of the GitHub repository arjitbhatnagar3-hue/meta-agent-learning, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-09-30). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/arjitbhatnagar3-hue/meta-agent-learning
- Repository: https://github.com/arjitbhatnagar3-hue/meta-agent-learning

## Overview

Meta AgentX accepts business-analysis requests, gathers reusable profile and document context, then runs durable, multi-agent investigations that produce reviewed, evidence-backed reports. The API serves the browser and manages tasks and business data; a separate worker claims queued tasks, plans and executes specialist work, and persists outcomes. Document storage, read-only business-system connectors, and selectable LLM providers support this workflow. The README is empty, so capability coverage relies on the tree and source index; unsampled modules are represented only where their roles are explicit.

Main entry points: [`backend/api.py`](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/api.py) (HTTP API), [`backend/worker.py`](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/worker.py) (Durable worker), [`backend/agents/registry.py`](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/agents/registry.py) (Agent registry), [`backend/agents/specialist.py`](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/agents/specialist.py) (Specialist agents)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_product["User experience"]
  node_browser["Web interface<br/>[app.js]"]
end

subgraph group_api["API and business data"]
  node_api["HTTP API<br/>[api.py]"]
  node_taskdata["Task and event repository<br/>[repository.py]"]
  node_database[("Application database<br/>[db.py]")]
  node_business["Business context<br/>[repository.py]"]
  node_documents["Document storage<br/>[storage.py]"]
end

subgraph group_execution["Task execution"]
  node_worker["Durable worker<br/>[worker.py]"]
  node_processor["Task processor<br/>[processor.py]"]
  node_workflow["Dependency workflow<br/>[workflow.py]"]
end

subgraph group_agents["Agent runtime"]
  node_manager["Manager planner<br/>[meta.py]"]
  node_registry["Agent registry<br/>[registry.py]"]
  node_specialist["Specialist agents<br/>[specialist.py]"]
  node_contracts["Agent contracts<br/>[models.py]"]
  node_toolregistry["Scoped tool registry<br/>[base.py]"]
end

subgraph group_services["Providers and tools"]
  node_llmcontract["LLM contract<br/>[base.py]"]
  node_fake["Fake LLM<br/>[fake.py]"]
  node_huggingface["Hugging Face provider<br/>[huggingface.py]"]
  node_connectors["CRM and billing tools<br/>[integrations.py]"]
  node_integrations["Business-system adapters<br/>[integrations.py]"]
  node_enterprise["Enterprise tool adapters<br/>[enterprise.py]"]
  node_settings["Runtime settings<br/>[config.py]"]
end

node_user(("Business user"))
node_objectstore[("Supabase Storage")]

node_user -->|"uses"| node_browser
node_browser -.->|"sends requests"| node_api
node_api -->|"manages tasks"| node_taskdata
node_api -->|"manages context"| node_business
node_api -->|"uploads and downloads"| node_documents
node_taskdata -->|"reads and writes"| node_database
node_business -->|"persists context"| node_database
node_documents -->|"uploads and retrieves"| node_objectstore
node_api -->|"lists agents"| node_registry
node_api -->|"loads configuration"| node_settings
node_worker -->|"claims and heartbeats"| node_taskdata
node_worker -->|"builds agents"| node_registry
node_worker -->|"constructs planner"| node_manager
node_worker -->|"constructs engine"| node_workflow
node_worker -->|"processes claimed tasks"| node_processor
node_processor -->|"loads and updates"| node_taskdata
node_processor -->|"requests plan"| node_manager
node_processor -->|"executes plan"| node_workflow
node_manager -->|"requests completion"| node_llmcontract
node_manager -->|"validates plan"| node_contracts
node_workflow -->|"runs assigned work"| node_specialist
node_specialist -->|"executes tools"| node_toolregistry
node_specialist -->|"requests completion"| node_llmcontract
node_specialist -->|"produces results"| node_contracts
node_registry -->|"creates specialists"| node_specialist
node_registry -->|"registers configured tools"| node_connectors
node_registry -->|"checks credentials"| node_settings
node_connectors -->|"calls adapters"| node_integrations
node_integrations -->|"uses credentials"| node_settings
node_toolregistry -.->|"supports tool contract"| node_enterprise
node_fake -->|"implements"| node_llmcontract
node_huggingface -->|"implements"| node_llmcontract
node_huggingface -->|"reads inference settings"| node_settings

click node_browser "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/frontend/app.js"
click node_api "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/api.py"
click node_taskdata "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/repository.py"
click node_database "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/db.py"
click node_business "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/repository.py"
click node_documents "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/storage.py"
click node_worker "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/worker.py"
click node_processor "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/processor.py"
click node_workflow "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/workflow.py"
click node_manager "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/agents/meta.py"
click node_registry "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/agents/registry.py"
click node_specialist "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/agents/specialist.py"
click node_contracts "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/models.py"
click node_toolregistry "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/tools/base.py"
click node_llmcontract "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/llm/base.py"
click node_fake "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/llm/fake.py"
click node_huggingface "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/llm/huggingface.py"
click node_connectors "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/tools/integrations.py"
click node_integrations "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/integrations.py"
click node_enterprise "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/tools/enterprise.py"
click node_settings "https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/config.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_browser,node_user toneBlue
class node_api,node_taskdata,node_database,node_business,node_documents,node_objectstore toneAmber
class node_worker,node_processor,node_workflow toneMint
class node_manager,node_registry,node_specialist,node_contracts,node_toolregistry toneRose
class node_llmcontract,node_fake,node_huggingface,node_connectors,node_integrations,node_enterprise,node_settings toneIndigo
```

## Components

### User experience

- [Web interface](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/frontend/app.js): `frontend/app.js`

### API and business data

- [HTTP API](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/api.py): `backend/api.py`
- [Task and event repository](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/repository.py): `backend/repository.py`
- [Application database](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/db.py): `backend/db.py`
- [Business context](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/repository.py): `backend/repository.py`
- [Document storage](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/storage.py): `backend/storage.py`

### Task execution

- [Durable worker](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/worker.py): `backend/worker.py`
- [Task processor](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/processor.py): `backend/processor.py`
- [Dependency workflow](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/workflow.py): `backend/workflow.py`

### Agent runtime

- [Manager planner](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/agents/meta.py): `backend/agents/meta.py`
- [Agent registry](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/agents/registry.py): `backend/agents/registry.py`
- [Specialist agents](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/agents/specialist.py): `backend/agents/specialist.py`
- [Agent contracts](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/models.py): `backend/models.py`
- [Scoped tool registry](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/tools/base.py): `backend/tools/base.py`

### Providers and tools

- [LLM contract](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/llm/base.py): `backend/llm/base.py`
- [Fake LLM](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/llm/fake.py): `backend/llm/fake.py`
- [Hugging Face provider](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/llm/huggingface.py): `backend/llm/huggingface.py`
- [CRM and billing tools](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/tools/integrations.py): `backend/tools/integrations.py`
- [Business-system adapters](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/integrations.py): `backend/integrations.py`
- [Enterprise tool adapters](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/tools/enterprise.py): `backend/tools/enterprise.py`
- [Runtime settings](https://github.com/arjitbhatnagar3-hue/meta-agent-learning/blob/main/backend/config.py): `backend/config.py`

### Other components

- **Business user**
- **Supabase Storage**

## Connections

- Business user → Web interface: uses
- Web interface → HTTP API: sends requests
- HTTP API → Task and event repository: manages tasks (evidence: `backend/api.py`)
- HTTP API → Business context: manages context (evidence: `backend/api.py`)
- HTTP API → Document storage: uploads and downloads (evidence: `backend/api.py`)
- Task and event repository → Application database: reads and writes (evidence: `backend/repository.py`)
- Business context → Application database: persists context (evidence: `backend/repository.py`)
- Document storage → Supabase Storage: uploads and retrieves (evidence: `backend/storage.py`)
- HTTP API → Agent registry: lists agents (evidence: `backend/api.py`)
- HTTP API → Runtime settings: loads configuration (evidence: `backend/api.py`)
- Durable worker → Task and event repository: claims and heartbeats (evidence: `backend/worker.py`)
- Durable worker → Agent registry: builds agents (evidence: `backend/worker.py`)
- Durable worker → Manager planner: constructs planner (evidence: `backend/worker.py`)
- Durable worker → Dependency workflow: constructs engine (evidence: `backend/worker.py`)
- Durable worker → Task processor: processes claimed tasks (evidence: `backend/worker.py`)
- Task processor → Task and event repository: loads and updates (evidence: `backend/processor.py`)
- Task processor → Manager planner: requests plan (evidence: `backend/processor.py`)
- Task processor → Dependency workflow: executes plan (evidence: `backend/processor.py`)
- Manager planner → LLM contract: requests completion (evidence: `backend/agents/meta.py`)
- Manager planner → Agent contracts: validates plan (evidence: `backend/agents/meta.py`)
- Dependency workflow → Specialist agents: runs assigned work (evidence: `backend/workflow.py`)
- Specialist agents → Scoped tool registry: executes tools (evidence: `backend/agents/specialist.py`)
- Specialist agents → LLM contract: requests completion (evidence: `backend/agents/specialist.py`)
- Specialist agents → Agent contracts: produces results (evidence: `backend/agents/specialist.py`)
- Agent registry → Specialist agents: creates specialists (evidence: `backend/agents/registry.py`)
- Agent registry → CRM and billing tools: registers configured tools (evidence: `backend/agents/registry.py`)
- Agent registry → Runtime settings: checks credentials (evidence: `backend/agents/registry.py`)
- CRM and billing tools → Business-system adapters: calls adapters (evidence: `backend/tools/integrations.py`)
- Business-system adapters → Runtime settings: uses credentials (evidence: `backend/integrations.py`)
- Scoped tool registry → Enterprise tool adapters: supports tool contract (evidence: `backend/tools/enterprise.py`)
- Fake LLM → LLM contract: implements (evidence: `backend/llm/fake.py`)
- Hugging Face provider → LLM contract: implements (evidence: `backend/llm/huggingface.py`)
- Hugging Face provider → Runtime settings: reads inference settings (evidence: `backend/llm/huggingface.py`)