# mukulsinghal001/multi-agent-workflow architecture

> Architecture diagram of the GitHub repository mukulsinghal001/multi-agent-workflow, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-20). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/mukulsinghal001/multi-agent-workflow
- Repository: https://github.com/mukulsinghal001/multi-agent-workflow

## Overview

This repository is a public-facing demo of a multi-agent business-analysis framework, not the framework’s implementation: Gemini CLI performs orchestration. The documented user workflow starts with a business goal and agent configuration, proceeds through hypothesis generation, analysis, research, reporting, and strategic review, then iterates as needed. The repository supplies generic data and guardrails, one generated analysis script, and sample run artifacts. The README documents the stages and CLI invocation; source inspection is limited to the analysis script, so internal agent dispatch and data handoffs are not shown.

Main entry points: [`run/analysis_code.py`](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/run/analysis_code.py) (Data Analysis), [`agents_demo.md`](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/agents_demo.md) (Report Writing), [`task_demo.txt`](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/task_demo.txt) (Business Goal), [`data_repository_demo.md`](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/data_repository_demo.md) (Generic Data)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_inputs["Demo Inputs"]
  node_goal["Business Goal<br/>[task_demo.txt]"]
  node_agent_personas["Agent Personas<br/>[agents_demo.md]"]
  node_data_repository["Generic Data"]
  node_guardrails["Guardrails<br/>[guardrails_demo.md]"]
  node_cli_config["Agent Config<br/>[cli_config.yaml]"]
end

subgraph group_workflow["Analysis Workflow"]
  node_hypothesis["Hypothesis Generation<br/>[agents_demo.md]"]
  node_analysis["Data Analysis<br/>[analysis_code.py]"]
  node_research["Market Research<br/>[agents_demo.md]"]
  node_strategic_review["Strategic Review<br/>[agents_demo.md]"]
  node_report_writer["Report Writing<br/>[agents_demo.md]"]
end

subgraph group_artifacts["Run Artifacts"]
  node_report["Sample Report<br/>[report.md]"]
  node_transcript["Sample Transcript<br/>[transcript.md]"]
end

node_business_user(("Business User"))
node_gemini_cli{{"Gemini CLI"}}

node_business_user -->|"sets goal"| node_goal
node_goal -->|"provides task"| node_gemini_cli
node_cli_config -->|"configures agents"| node_gemini_cli
node_gemini_cli -->|"starts workflow"| node_hypothesis
node_agent_personas -->|"defines roles"| node_hypothesis
node_hypothesis -->|"tests hypotheses"| node_analysis
node_data_repository -->|"supplies schemas"| node_analysis
node_analysis -.->|"informs research"| node_research
node_research -->|"supports synthesis"| node_report_writer
node_analysis -->|"supplies insights"| node_report_writer
node_guardrails -.->|"constrains operation"| node_gemini_cli
node_report_writer -->|"produces report"| node_report
node_gemini_cli -.->|"records run"| node_transcript
node_report_writer -->|"submits report"| node_strategic_review
node_strategic_review -->|"requests refinement"| node_hypothesis

click node_goal "https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/task_demo.txt"
click node_agent_personas "https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/agents_demo.md"
click node_data_repository "https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/data_repository_demo.md"
click node_guardrails "https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/guardrails_demo.md"
click node_cli_config "https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/cli_config.yaml"
click node_hypothesis "https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/agents_demo.md"
click node_analysis "https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/run/analysis_code.py"
click node_research "https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/agents_demo.md"
click node_strategic_review "https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/agents_demo.md"
click node_report_writer "https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/agents_demo.md"
click node_report "https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/run/report.md"
click node_transcript "https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/run/transcript.md"

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_goal,node_agent_personas,node_data_repository,node_guardrails,node_cli_config,node_business_user toneBlue
class node_hypothesis,node_analysis,node_research,node_strategic_review,node_report_writer toneAmber
class node_report,node_transcript toneMint
class node_gemini_cli toneIndigo
```

## Components

### Demo Inputs

- [Business Goal](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/task_demo.txt): `task_demo.txt`
- [Agent Personas](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/agents_demo.md): `agents_demo.md`
- [Generic Data](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/data_repository_demo.md): `data_repository_demo.md`
- [Guardrails](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/guardrails_demo.md): `guardrails_demo.md`
- [Agent Config](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/cli_config.yaml): `cli_config.yaml`

### Analysis Workflow

- [Hypothesis Generation](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/agents_demo.md): `agents_demo.md`
- [Data Analysis](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/run/analysis_code.py): `run/analysis_code.py`
- [Market Research](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/agents_demo.md): `agents_demo.md`
- [Strategic Review](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/agents_demo.md): `agents_demo.md`
- [Report Writing](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/agents_demo.md): `agents_demo.md`

### Run Artifacts

- [Sample Report](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/run/report.md): `run/report.md`
- [Sample Transcript](https://github.com/mukulsinghal001/multi-agent-workflow/blob/main/run/transcript.md): `run/transcript.md`

### Other components

- **Business User**
- **Gemini CLI**

## Connections

- Business User → Business Goal: sets goal (evidence: `README.md`)
- Business Goal → Gemini CLI: provides task (evidence: `README.md`)
- Agent Config → Gemini CLI: configures agents (evidence: `README.md`)
- Gemini CLI → Hypothesis Generation: starts workflow (evidence: `README.md`)
- Agent Personas → Hypothesis Generation: defines roles (evidence: `README.md`)
- Hypothesis Generation → Data Analysis: tests hypotheses (evidence: `README.md`)
- Generic Data → Data Analysis: supplies schemas (evidence: `README.md`)
- Data Analysis → Market Research: informs research
- Market Research → Report Writing: supports synthesis (evidence: `README.md`)
- Data Analysis → Report Writing: supplies insights (evidence: `README.md`)
- Guardrails → Gemini CLI: constrains operation
- Report Writing → Sample Report: produces report (evidence: `README.md`)
- Gemini CLI → Sample Transcript: records run
- Report Writing → Strategic Review: submits report (evidence: `README.md`)
- Strategic Review → Hypothesis Generation: requests refinement (evidence: `README.md`)
