Tutorials
Learn the core features of AllvIa through real-world business scenarios. Organized by difficulty level, from basic to advanced.
Basic
Natural Language OS Control (surf command)
| Time to Complete | ~10 minutes |
| Difficulty | Basic |
| Required Connectors | None (Core feature) |
| Learning Objectives | surf command, NL pipeline, multi-app scenarios |
Use the surf command, a core feature of AllvIa, to control your OS directly with natural language.
Step 1 — Basic surf Command
Control apps with natural language. The intent_router analyzes intent, and plan_builder generates an execution plan:
> surf "Open Safari and search for weather on Google"
Intent: web_search | Slots: {app: Safari, query: weather}
Plan: Open Safari → Navigate → Search
[OK] Safari execution complete (applescript.rs)
[OK] Google search complete (browser_automation.rs)
> surf "Open Notes, create a new note, and type Today's To-Do List"
Intent: note_create | Slots: {app: Notes, content: Today's To-Do List}
[OK] Notes.app launched
[OK] Shortcut: Cmd+N (New note created)
[OK] Type: "Today's To-Do List" (keyboard.type) Step 2 — Multi-App Scenarios
Complex workflows connecting multiple apps can be executed with a single line:
> surf "Search Apple stock price in Safari, calculate value of 100 shares in Calculator, then save result in Notes"
Plan: 4 steps
1. Open Safari → Search "Apple stock price"
2. Vision: Read price from screen
3. Open Calculator → Multiply × 100
4. Open Notes → Save result
[OK] 4/4 steps completed Step 3 — Reality Check
Before execution, reality_check.rs scans the list of installed apps and automatically corrects similar names. For example, "Excel" will be automatically matched to "Microsoft Excel".
Automated Weekly Report Generation
| Time to Complete | ~30 minutes |
| Difficulty | Intermediate |
| Required Connectors | Google Sheets, Slack |
| Learning Objectives | Schedule triggers, data collection, report formatting |
Collect KPIs from multiple data sources to automatically generate and share reports to team channels every Monday morning.
Step 1 — Configure Schedule Trigger
> build_workflow "Collect KPIs from Google Sheets and send weekly report to Slack every Monday at 9 AM"
Analyzing with LLM...
Generating n8n workflow:
├─ Trigger: Schedule — Every Monday 09:00 KST
├─ Action: Google Sheets — Read KPI Data
├─ Action: Format — Generate Markdown Report
└─ Action: Slack — Post to #team-weekly
workflow_weekly_report.json generation complete Step 2 — Connect Data Sources
Configure the Google Sheets connector in the n8n dashboard and link the sheet containing KPI data. You can consolidate data from multiple sheets into a single report.
Step 3 — Set Report Format
Configure the data fields and visualization options to be included in the report:
- Weekly revenue change (% compared to previous week)
- New user count and churn rate
- Achievement rate relative to target
- Key issues and action items
Advanced
Telegram Remote Control Setup
| Time to Complete | ~30 minutes |
| Difficulty | Advanced |
| Required Connectors | Telegram Bot API |
| Learning Objectives | Remote command execution, security policies, result reporting |
Integrate a Telegram Bot to remotely control your computer from your phone while away. telegram.rs (12KB) handles the Bot API integration.
Step 1 — Create Telegram Bot
Create a Telegram Bot via BotFather and obtain a token:
# Add Telegram settings to .env file
TELEGRAM_BOT_TOKEN=your_bot_token
TELEGRAM_CHAT_ID=your_chat_id Step 2 — Execute Remote Commands
When you send a message from Telegram, AllvIa executes the command and reports the result:
# Sent from Telegram:
Phone: "Check git status on computer"
# AllvIa Processing:
Intent: shell_exec | Security: Safe
> git status
# Telegram Report:
✅ Execution Complete
On branch main
Your branch is up to date with 'origin/main'.
nothing to commit, working tree clean Step 3 — Verify Security Policies
The same 5-layer security applies to remote commands. Critical-level commands (rm, sudo, etc.) are blocked even remotely, and Write Lock can only be released locally.
Custom Automation Based on Repeating Patterns
| Time to Complete | ~20 minutes |
| Difficulty | Intermediate |
| Required Connectors | None (Core feature) |
| Learning Objectives | routine analysis, recommend usage, applying automation |
Automate personal repetitive tasks using AllvIa's core behavioral pattern analysis feature.
Step 1 — Check Data Collection Status
Check the amount of currently collected data with the status command:
> status
AllvIa Agent Status:
Collector: running (port 8080)
Database: steer.db (42.3 MB)
Events: 128,450 events (last 7 days)
Sessions: 2,340 sessions built
Routines: 18 patterns detected
Write Lock: enabled Step 2 — Verify Pattern Analysis & Recommendations
Review detected patterns with routine and get AI-based automation suggestions with recommend. Selecting a suggested item will immediately generate a workflow.
Step 3 — Deploy & Monitor Workflow
Register the generated workflow in n8n and monitor execution results. The Self-Healing Guardian ensures process stability.
Next Steps
- API Reference — Programmatic integration with the local core agent pipeline
- FAQ — Frequently asked questions and troubleshooting
- Changelog — Track latest feature updates
New tutorials are posted continuously. Check for new content in Updates.