Skip to main content

​
Quickstart

Get Risoluto running and process your first issue in under 10 minutes.

​
Prerequisites

RequirementDetails
Node.jsv22 or newer
DockerDocker Engine running (docker info should succeed)
pnpmPackage manager (npm install -g pnpm)
LinearA Linear account with at least one project

​
Install & Start

Clone and build

git clone https://github.com/OmerFarukOruc/risoluto.git
cd risoluto
pnpm install && pnpm run build

Build the sandbox image

This creates the Docker image that isolated AI agents run inside:
bash bin/build-sandbox.sh
The sandbox image is based on Ubuntu 24.04 with Node.js 22 and the Codex CLI pre-installed. It provides process isolation, resource limits, and security hardening out of the box.

Start Risoluto

node dist/cli/index.js --port 4000
Open http://127.0.0.1:4000 — the setup wizard opens automatically on first run.

​
Setup Wizard

The wizard walks you through four steps:

Protect secrets

Generates an AES-256 encryption master key that protects all stored credentials on disk. This key never leaves your machine.

Connect Linear

Paste your Linear API key. The wizard validates it, lists your projects, and lets you select the one to monitor.

Add OpenAI

Choose one of three authentication methods:
Paste an sk-... API key directly. Risoluto validates and encrypts it.

Add GitHub (optional)

Paste a GitHub PAT to enable automatic PR creation when agents complete work.

​
Verify It Works

Create a test issue in your Linear project: Title: 
SMOKE: create workspace proof file
Description: 
Create `RISOLUTO_SMOKE_RESULT.md` in the workspace with:
- the issue identifier
- the current UTC timestamp
- the output of `pwd` and `ls -la`
- whether the workspace looks empty or repo-backed
Set the issue to In Progress. Within one poll cycle (default: 15 seconds), Risoluto picks it up automatically.
How to confirm it’s working: The issue appears on the dashboard at http://127.0.0.1:4000 with real-time progress indicators. You can also check the API directly.

​
Check the Results

Dashboard

The issue appears on the board at http://127.0.0.1:4000 with real-time progress.

API

curl -s http://127.0.0.1:4000/api/v1/state | jq
When the agent finishes, inspect the workspace: 
cat ../risoluto-workspaces/<ISSUE_ID>/RISOLUTO_SMOKE_RESULT.md

​
What’s Next

How It Works

Understand the full architecture — polling, workspaces, sandboxes, and delivery.

Docker Deployment

Deploy with Docker Compose for persistent, unattended operation.

Configuration

Customize models, timeouts, retries, and workspace strategies.

Network Security

Secure the API with write tokens, bind address controls, and TLS.
Last modified on March 31, 2026