Installation
Tessera is distributed as a Python package and can be run directly with uvx (recommended) or installed globally.
Requirements
- Python: 3.13 or higher
- uv: Package manager (recommended)
- API Keys: OpenAI, Anthropic, or compatible provider
Installation Methods
Tessera is published to PyPI as tessera-agents. You can either run it directly with uvx or install it globally.
Option 1: Run Directly (Recommended)
Use uvx to run Tessera without installation:
Benefits: - No installation needed - Always uses latest version - Isolated execution environment
Note: When using uvx, you must include tessera-agents in the command (e.g., uvx tessera-agents <command>).
Option 2: Global Installation
Install once, then use the shorter tessera command:
# Install the package
uv tool install tessera-agents
# Now use the 'tessera' command directly
tessera init
tessera main "Build a FastAPI app"
Benefits: - Shorter commands after installation - Faster execution (no download each time) - Works offline after installation
Note: The PyPI package name is tessera-agents, but after installation, you use the tessera command.
What Gets Created
Both installation methods create the same configuration:
~/.config/tessera/
├── config.yaml # Main configuration
└── prompts/ # Agent system prompts
├── supervisor.md
└── ...
~/.cache/tessera/
├── metrics.db # Task history and costs
└── otel/
└── traces.jsonl # LLM call traces
Configuration
On first run, Tessera creates:
~/.config/tessera/
├── config.yaml # Main configuration
└── prompts/ # Agent system prompts
├── supervisor.md
└── ...
~/.cache/tessera/
├── metrics.db # Task history and costs
└── otel/
└── traces.jsonl # LLM call traces
API Keys
Tessera needs LLM API keys. Set them via environment variables:
# Option 1: .env file (in your project directory)
echo "OPENAI_API_KEY=sk-your-key-here" >> .env
# Option 2: Export directly
export OPENAI_API_KEY=sk-your-key-here
export ANTHROPIC_API_KEY=sk-ant-your-key-here
Verify Installation
Expected output: