This guide shows how to install AI Native Lang (AINL) on your machine and run your first checks.
If you use an AI agent (like OpenClaw, Claude, ChatGPT, Cursor, or others), you can also ask it to do most of this setup for you.
Quick option: ask your AI agent to install AINL
AINL works with any AI agent, not just OpenClaw.
Copy and paste this into your agent:
Install AINL from https://github.com/sbhooley/ainativelang and guide me through setup in simple steps.
You can follow the manual steps below if you prefer full control.
Prerequisites
- Python 3.10+
git- A local terminal (macOS/Linux/Windows)
1) Clone the repo
git clone https://github.com/sbhooley/ainativelang.git
cd ainativelang
2) Create and activate a virtual environment
python -m venv .venv
source .venv/bin/activate # on Windows: .venv\\Scripts\\activate
python -m pip install --upgrade pip
3) Install AINL (editable)
pip install -e ".[dev,web]"
4) Validate the toolchain
Validate an example workflow in strict mode:
python scripts/validate_ainl.py examples/hello.ainl --strict
Inspect the compiled graph IR:
ainl-validate examples/hello.ainl --strict --emit ir
5) Run the test security profiles
Run the core test profile:
python scripts/run_test_profiles.py --profile core
6) Start the runner service (optional)
This is useful if you want to integrate AINL via an HTTP API surface:
python scripts/runtime_runner_service.py
Then you can call:
GET http://localhost:8770/capabilitiesPOST http://localhost:8770/run
Need a hand? We offer expert help.
AINL is a powerful system, and sometimes the fastest way to move forward is to have someone who knows it well in your corner.
Whether you’re trying to get a clean install working, you’ve hit a validation error you can’t explain, or you’re planning a production deployment and want to get the architecture right — we’re available for paid engagements:
- Installation & setup — We’ll get AINL running in your local environment, CI pipeline, or container, configured for your stack and ready to go.
- Troubleshooting — Strict validation failures, adapter issues, runtime errors, profile configuration — we’ll find the problem and fix it with you.
- Architecture consultations — Designing a real agent system? We can help you think through adapter design, MCP host integration, security profiles, and the full compile-and-run story before you build.
Next steps
- Read the official docs at
/docs - Browse more guides at
/learn - See practical applications at
/case-studies - If you still need OpenClaw setup first, use How to Install & Setup OpenClaw