Getting Started

---

title: "Getting Started" tier: "1" status: "verified" version: "1.2.0" last-verified: "2026-06-17" commit: "b1c347b1" network-badge: "required" risk-tags:

• "RISK: HIGH"
• "RISK: MEDIUM"

---

🚀 Getting Started

☁️ REQUIRES NETWORK

This guide walks you through booting your Tadpole OS local node, accessing the dashboard for the first time, and running your first virtual employee project.

---

1. Prerequisites

Before starting, ensure you have the following installed on your office computer:

Requirement	Minimum Version	Purpose
Node.js	v18+	Frontend build tooling
Rust	1.75+	Backend engine compilation
Python	3.10+	Execution scripts and parity checks
Ollama (optional)	Latest	Local LLM inference (fully offline)

Tip

For a fully local, zero-network setup, install Ollama and download a model such as llama3:8b. This allows all AI processing to remain on your hardware.

→ See Installation for detailed compilation and dependency steps.

---

2. Starting the Engine

2.1 Start the Backend (Rust Engine)

Navigate to the server-rs/ directory and run:

cargo run --release

The engine binds to 127.0.0.1:8000 by default (configurable via PORT [RISK: LOW] and BIND_ADDRESS [RISK: HIGH]).

Note

Runtime Governance Anchor: All persistence paths resolve from AppState.base_dir, ensuring consistent folder roots across local-dev and server-mounted office networks.

2.2 Start the Frontend (Vite Dashboard)

In a separate terminal, from the project root:

npm run dev

The dashboard opens at http://localhost:5173 by default.

2.3 Verify Connection

Once both services are running, open your browser to http://localhost:5173. The Engine Status indicator in the PageHeader should show 🟢 ONLINE, confirming the frontend is connected to the backend via WebSocket.

---

3. First-Time Dashboard Tour

The Tadpole OS dashboard is a multi-tab, multi-context interface:

1. Dashboard Tab: Your central operations hub showing engine health, active agents, and system metrics.
2. Org Chart Tab: A visual force-graph showing your virtual employee hierarchy.
3. Standups Tab: Voice-driven team coordination portal.
4. Workspaces Tab: Manage project folders, shared documents, and file assets.
5. Docs Tab: Access to embedded documentation and knowledge bases.
6. Settings Tab: Configure providers, profiles, and system parameters.

→ See Dashboard-Guide for a complete breakdown of every interactive element.

3.1 Tab Navigation Anchors

• [Tab] Tab Item
  • Default state: Active on first tab (Dashboard).
  • Visible location: Multi-Tab Bar at the top of the viewport.
  • Observable side effect: Focuses the selected tab area and preserves navigation states.
  • Keyboard shortcut: [Shortcut: 1-6] to route to tabs 1-6.
  • Missing in build: Always present.

---

4. Running Your First Project

Step 1: Configure a Provider

Navigate to Settings → Model Manager and unlock the Glossary#secure-credentials-vault by entering a master passphrase. See LLM-Providers for the vault visual reference anchors. Add at least one LLM provider:

• Local (Ollama): Set OLLAMA_HOST [RISK: MEDIUM] to http://localhost:11434 (default).
• Cloud (OpenAI, Anthropic, etc.): Enter the API key for your chosen provider.

Important

Cloud providers require a network connection. Prompt payloads and decrypted API tokens are sent directly to the provider endpoint. See LLM-Providers for the complete list of supported providers and their data paths.

Step 2: Create a Virtual Employee

Navigate to Settings → Agent Configuration and create a new agent profile. See Virtual-Employees for Agent Configuration Panel visual reference anchors:

• Name: e.g., "Support Analyst"
• Model: Select your configured LLM (e.g., llama3:8b via Ollama).
• Budget: Set a spending cap (default: $1.00 USD per agent, configurable via DEFAULT_AGENT_BUDGET_USD [RISK: MEDIUM]).

→ See Virtual-Employees for advanced profile tuning.

Step 3: Create a Project Workspace

Navigate to Workspaces and create a new project folder (see Projects-&-Missions for project creation wizard visual reference anchors):

• Name: e.g., "Customer Feedback Q2"
• Folder Path: Select a local directory containing your business documents (e.g., reviews.csv).
• Assign Agent: Attach your "Support Analyst" to this workspace.

Step 4: Issue a Directive

Open the Chat-&-Voice panel and type:

Analyze reviews.csv, group them into categories, and write a summary report.txt in the same directory.

Step 5: Approve Actions

The agent will request permission to read files. Approve the read_file action in the Oversight Queue on the Approvals-&-Quotas page. See Approvals-&-Quotas for Oversight Queue visual reference anchors.

Step 6: Review Output

Once the agent completes, review the generated draft on the Workspaces page and click [Button] Approve Draft (see Projects-&-Missions for draft approval UI anchors) to merge it into your project folder.

---

5. Quick Reference: Keyboard Shortcuts

Shortcut	Action
[Shortcut: Ctrl+K] or [Shortcut: Ctrl+/]	Toggle Command Palette
[Shortcut: 1-6]	Quick-navigate to tabs (Dashboard, Org Chart, Standups, Workspaces, Docs, Settings)

Note

Keyboard shortcuts are active only when no input fields are focused. Cited in useLayoutNavigation.ts:L51-83.

---

6. What's Next?

Goal	Go To
Explore the full dashboard layout	Dashboard-Guide
Learn chat and voice commands	Chat-&-Voice
Set up multiple projects	Projects-&-Missions
Follow real SMB workflows	Daily-Workflows
Configure environment variables	Configuration
Understand security controls	Vault-&-Credentials

---

Complete Lexicon: For the authoritative technical breakdown, see the main repository GLOSSARY.md. Every [[Glossary#term|term]] link on this page resolves to an entry there.