# atomslight/nordavo_project_support_agent architecture

> Architecture diagram of the GitHub repository atomslight/nordavo_project_support_agent, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-21). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/atomslight/nordavo_project_support_agent
- Repository: https://github.com/atomslight/nordavo_project_support_agent

## Overview

Nordavo is a CLI support-ticket agent. The user runs analysis or decision commands; workflows fetch tickets, redact sensitive text, classify language, and—when deciding—use the FAQ to prepare a validated reply, handoff, or closure. Decisions are persisted locally; sending is a separate explicit operation that posts approved saved actions to the Helpdesk. The graph also shows provider fallback, local state and audit logging. Relationships follow the supplied excerpts, SOURCE INDEX references, and README; unsampled subsystem internals are not inferred.

Main entry points: [`src/nordavo/decision_workflow.py`](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/decision_workflow.py) (Decision and send workflow), [`src/nordavo/cli.py`](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/cli.py) (Command interface), [`src/nordavo/ai/decision.py`](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/ai/decision.py) (Decision classifier), [`src/nordavo/ai/language.py`](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/ai/language.py) (Language classifier)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_entry["CLI and configuration"]
  node_cli["Command interface<br/>[cli.py]"]
  node_config["Settings<br/>[config.py]"]
end

subgraph group_processing["Ticket processing"]
  node_analysis["Language analysis workflow<br/>[workflow.py]"]
  node_decide["Decision and send workflow"]
  node_helpdesk_client["Helpdesk client<br/>[helpdesk.py]"]
  node_ticket_models["Ticket and decision models<br/>[models.py]"]
end

subgraph group_ai["AI and safety"]
  node_safety["Redaction and validation<br/>[safety.py]"]
  node_language["Language classifier<br/>[language.py]"]
  node_decision["Decision classifier<br/>[decision.py]"]
  node_prompts["Prompt rules<br/>[prompts.py]"]
end

subgraph group_persistence["State and audit"]
  node_ticket_state[("Ticket state store<br/>[state.py]")]
  node_decision_state[("Decision ledger<br/>[decision_state.py]")]
  node_audit["Run audit log<br/>[run_log.py]"]
end

node_operator(("CLI user"))
node_helpdesk["Helpdesk API"]
node_model_services["Gemini and Groq"]
node_faq["Support FAQ<br/>[faq.md]"]

node_operator -->|"runs commands"| node_cli
node_cli -->|"loads settings"| node_config
node_cli -->|"starts analysis"| node_analysis
node_cli -->|"starts decisions"| node_decide
node_cli -->|"builds client"| node_helpdesk_client
node_cli -->|"builds analyzer"| node_language
node_cli -->|"builds analyzer"| node_decision
node_cli -->|"creates audit"| node_audit
node_analysis -->|"fetches tickets"| node_helpdesk_client
node_analysis -->|"redacts tickets"| node_safety
node_analysis -->|"analyzes text"| node_language
node_analysis -->|"saves results"| node_ticket_state
node_decide -->|"fetches tickets"| node_helpdesk_client
node_decide -->|"redacts tickets"| node_safety
node_decide -->|"analyzes language"| node_language
node_decide -->|"requests decisions"| node_decision
node_decide -->|"reads and updates"| node_ticket_state
node_decide -->|"saves decisions"| node_decision_state
node_decide -->|"records events"| node_audit
node_decide -->|"reads FAQ"| node_faq
node_decide -->|"sends approved actions"| node_helpdesk_client
node_helpdesk_client -->|"requests tickets and posts"| node_helpdesk
node_language -->|"classifies with fallback"| node_model_services
node_decision -->|"decides with fallback"| node_model_services
node_decision -->|"uses decision rules"| node_prompts
node_language -->|"uses language prompt"| node_prompts
node_decision -->|"validates output"| node_safety
node_language -->|"returns assessment"| node_ticket_models
node_decision -->|"returns decision"| node_ticket_models
node_helpdesk_client -->|"maps ticket data"| node_ticket_models
node_ticket_state -->|"stores ticket data"| node_ticket_models
node_decision_state -->|"stores decisions"| node_ticket_models

click node_cli "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/cli.py"
click node_config "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/be/config.py"
click node_analysis "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/workflow.py"
click node_decide "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/decision_workflow.py"
click node_helpdesk_client "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/be/helpdesk.py"
click node_safety "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/ai/safety.py"
click node_language "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/ai/language.py"
click node_decision "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/ai/decision.py"
click node_prompts "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/ai/prompts.py"
click node_ticket_models "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/models.py"
click node_faq "https://github.com/atomslight/nordavo_project_support_agent/blob/main/data/knowledgebase/faq.md"
click node_ticket_state "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/be/state.py"
click node_decision_state "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/be/decision_state.py"
click node_audit "https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/be/run_log.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_cli,node_config,node_operator toneBlue
class node_analysis,node_decide,node_helpdesk_client,node_ticket_models toneAmber
class node_safety,node_language,node_decision,node_prompts,node_helpdesk toneMint
class node_ticket_state,node_decision_state,node_audit toneRose
class node_model_services toneIndigo
class node_faq toneTeal
```

## Components

### CLI and configuration

- [Command interface](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/cli.py): `src/nordavo/cli.py`
- [Settings](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/be/config.py): `src/nordavo/be/config.py`

### Ticket processing

- [Language analysis workflow](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/workflow.py): `src/nordavo/workflow.py`
- [Decision and send workflow](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/decision_workflow.py): `src/nordavo/decision_workflow.py`
- [Helpdesk client](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/be/helpdesk.py): `src/nordavo/be/helpdesk.py`
- [Ticket and decision models](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/models.py): `src/nordavo/models.py`

### AI and safety

- [Redaction and validation](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/ai/safety.py): `src/nordavo/ai/safety.py`
- [Language classifier](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/ai/language.py): `src/nordavo/ai/language.py`
- [Decision classifier](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/ai/decision.py): `src/nordavo/ai/decision.py`
- [Prompt rules](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/ai/prompts.py): `src/nordavo/ai/prompts.py`

### State and audit

- [Ticket state store](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/be/state.py): `src/nordavo/be/state.py`
- [Decision ledger](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/be/decision_state.py): `src/nordavo/be/decision_state.py`
- [Run audit log](https://github.com/atomslight/nordavo_project_support_agent/blob/main/src/nordavo/be/run_log.py): `src/nordavo/be/run_log.py`

### Other components

- **CLI user**
- **Helpdesk API**
- **Gemini and Groq**
- [Support FAQ](https://github.com/atomslight/nordavo_project_support_agent/blob/main/data/knowledgebase/faq.md): `data/knowledgebase/faq.md`

## Connections

- CLI user → Command interface: runs commands (evidence: `README.md`)
- Command interface → Settings: loads settings (evidence: `src/nordavo/cli.py`)
- Command interface → Language analysis workflow: starts analysis (evidence: `src/nordavo/cli.py`)
- Command interface → Decision and send workflow: starts decisions (evidence: `src/nordavo/cli.py`)
- Command interface → Helpdesk client: builds client (evidence: `src/nordavo/cli.py`)
- Command interface → Language classifier: builds analyzer (evidence: `src/nordavo/cli.py`)
- Command interface → Decision classifier: builds analyzer (evidence: `src/nordavo/cli.py`)
- Command interface → Run audit log: creates audit (evidence: `src/nordavo/cli.py`)
- Language analysis workflow → Helpdesk client: fetches tickets (evidence: `src/nordavo/workflow.py`)
- Language analysis workflow → Redaction and validation: redacts tickets (evidence: `src/nordavo/workflow.py`)
- Language analysis workflow → Language classifier: analyzes text (evidence: `src/nordavo/workflow.py`)
- Language analysis workflow → Ticket state store: saves results (evidence: `src/nordavo/workflow.py`)
- Decision and send workflow → Helpdesk client: fetches tickets (evidence: `src/nordavo/decision_workflow.py`)
- Decision and send workflow → Redaction and validation: redacts tickets (evidence: `src/nordavo/decision_workflow.py`)
- Decision and send workflow → Language classifier: analyzes language (evidence: `src/nordavo/decision_workflow.py`)
- Decision and send workflow → Decision classifier: requests decisions (evidence: `src/nordavo/decision_workflow.py`)
- Decision and send workflow → Ticket state store: reads and updates (evidence: `src/nordavo/decision_workflow.py`)
- Decision and send workflow → Decision ledger: saves decisions (evidence: `src/nordavo/decision_workflow.py`)
- Decision and send workflow → Run audit log: records events (evidence: `src/nordavo/decision_workflow.py`)
- Decision and send workflow → Support FAQ: reads FAQ (evidence: `README.md`)
- Decision and send workflow → Helpdesk client: sends approved actions (evidence: `src/nordavo/decision_workflow.py`)
- Helpdesk client → Helpdesk API: requests tickets and posts (evidence: `README.md`)
- Language classifier → Gemini and Groq: classifies with fallback (evidence: `src/nordavo/ai/language.py`)
- Decision classifier → Gemini and Groq: decides with fallback (evidence: `src/nordavo/ai/decision.py`)
- Decision classifier → Prompt rules: uses decision rules (evidence: `src/nordavo/ai/decision.py`)
- Language classifier → Prompt rules: uses language prompt (evidence: `src/nordavo/ai/language.py`)
- Decision classifier → Redaction and validation: validates output (evidence: `src/nordavo/ai/decision.py`)
- Language classifier → Ticket and decision models: returns assessment (evidence: `src/nordavo/ai/language.py`)
- Decision classifier → Ticket and decision models: returns decision (evidence: `src/nordavo/ai/decision.py`)
- Helpdesk client → Ticket and decision models: maps ticket data (evidence: `src/nordavo/be/helpdesk.py`)
- Ticket state store → Ticket and decision models: stores ticket data (evidence: `src/nordavo/be/state.py`)
- Decision ledger → Ticket and decision models: stores decisions (evidence: `src/nordavo/be/decision_state.py`)
