Quick Start
This tutorial walks you through using Tessera to generate your first project in under 10 minutes.
Installation
Choose your preferred method:
# Option 1: Run directly with uvx
uvx tessera-agents init
# Option 2: Install globally, then use shorter commands
uv tool install tessera-agents
For the rest of this guide, we'll use tessera commands (assumes you've installed globally). If using uvx, replace tessera with uvx tessera-agents.
Step 1: Initialize Configuration
Run the interactive setup wizard:
The wizard will ask: - Which LLM provider to use (OpenAI, Anthropic, etc.) - Your API key - Default agent configuration
This creates ~/.config/tessera/config.yaml.
Step 2: Generate Your First Project
Start Tessera in interactive mode:
Interactive Prompts
? What would you like to build?
> A simple Python CLI tool that fetches weather data
? Complexity level: Simple
? Interview mode (recommended for better results)?
> Yes
Interview Phase
The Interviewer agent will ask clarifying questions:
Q1/5: Which weather API should we use?
> OpenWeatherMap
Q2/5: What data should be displayed?
> Current temperature, conditions, and 5-day forecast
Q3/5: How should users provide location?
> City name as command-line argument
...
Planning Phase
The Supervisor agent generates a comprehensive plan:
✓ Plan generated: 8 tasks across 5 phases
✓ Agents required: 4 (code-architect, python-expert, test-engineer, tech-writer)
✓ Estimated cost: $1.20 | Time: ~45 minutes
Tasks:
1. Design CLI architecture
2. Implement weather API client
3. Create command-line interface
4. Add error handling
5. Write unit tests
6. Write integration tests
7. Create README documentation
8. Generate usage examples
? Approve this plan? [Y/n]:
Execution Phase
Starting execution...
Progress: ██████░░░░░░ 50% (4/8 tasks) | Cost: $0.62
ACTIVE AGENTS (2/3):
● python-expert [T2] Weather API client (3m)
● test-engineer [T5] Unit tests (1m)
COMPLETED:
✓ T1 CLI architecture design 2m12s $0.08
✓ T3 Command-line interface 5m32s $0.24
[Press 'P' to pause | 'Q' to quit]
Completion
✓ Project generation complete!
Generated:./weather-cli/
├── src/
│ ├── __init__.py
│ ├── client.py
│ └── cli.py
├── tests/
│ ├── test_client.py
│ └── test_cli.py
├── README.md
├── pyproject.toml
└── .gitignore
Summary:
- 8 files created
- 245 lines of code
- 92% test coverage
- All tests passing ✓
- Documentation complete ✓
Total cost: $1.18 | Duration: 42m
? Open project directory? [Y/n]:
Step 3: Review the Generated Project
The README will include: - Project description - Installation instructions - Usage examples - API documentation - Development guide
Step 4: Run the Generated Project
# Install dependencies
uv sync
# Run tests
uv run pytest
# Use the CLI
uv run python -m src.cli "San Francisco"
Next Steps
Customize Agents
Edit ~/.config/tessera/prompts/python-expert.md to customize how the agent writes code.
Adjust Configuration
Edit ~/.config/tessera/config.yaml to:
- Add more agents
- Configure cost limits
- Enable tools and plugins
- Set up Slack notifications