Keep control and keys in your environment
CLI, orchestration, safety policy, and reports run in your environment. Best for evaluating VOLY and individual work.
Install LocalYOUR FIRST MANAGED RUN
Run VOLY locally
Install the control plane, connect one provider, and begin with a dry-run. File changes are rolled back while the diff and run report remain available for review.
Time to first report≈ 10 minutes
Python3.10–3.14
SystemmacOS · Linux · WSL
BringGit + one provider key
CHOOSE YOUR RUNTIME
Local now or Cloud for the team
Share one surface without manual setup
Organizations, members, hosted runs, and shared spend history live in the team dashboard.
Open CloudLOCAL · 4 STEPS
From clone to a verified result
Run these commands in order from the directory where you want to keep the VOLY source.
01
Get the code
Clone the repository
Use a separate virtual environment so VOLY dependencies stay isolated from your projects.
git clone https://github.com/voly-codes/voly.git
cd voly
python3 -m venv .venv && source .venv/bin/activate
Your shell is now inside the active .venv environment.
02
Install
Add the CLI and Web UI
The editable install is convenient for core updates and includes the local dashboard dependencies.
pip install -e ".[ui,dev]"
voly --help
The voly command is available in the current shell.
03
Configure
Add a key and create the config
Copy the environment example, add at least one API key, and create voly.yaml. Agents cannot write to .env by default.
cp .env.example .env
# add a provider key to .env, then:
voly init && voly status
Status shows at least one available executor or provider.
04
Verify safely
Run your first dry-run
Point VOLY at a test repository. The agent completes the task, but file writes are rolled back and the diff remains in the report.
voly run "review README and suggest one improvement" --cwd ~/my-project --dry-run
The repository is unchanged; the result, touched files, and dry-run diff are saved.
AFTER THE RUN
Inspect evidence, not only the answer
Read the latest run in the CLI or start the local dashboard on port 7788.
voly runs list
voly runs show <task_id>
voly ui