Quickstart
This guide gets you from a fresh install to a running team in a few minutes.
Shortest path
# 1. Install prerequisites
node --version # need 20+
pnpm --version # need 10+
# 2. Clone and install
git clone https://github.com/777genius/agent-teams-ai.git
cd agent-teams-ai
pnpm install
# 3. Start the desktop app (default workflow)
pnpm dev
# 4. Verify a docs-only change
pnpm --dir landing docs:buildThe desktop Electron app (pnpm dev) is the primary target — do not use the browser/web dev server for normal development. The browser path lacks desktop IPC, terminal, provider auth, and team lifecycle behavior.
Before you begin
You need:
- A computer running macOS, Windows, or Linux
- (Recommended) A Git-tracked project — worktree isolation and diff review rely on Git
- (Optional) Provider access — runtime setup detects available providers from the UI, but some paths need existing auth (Anthropic, OpenAI, etc.)
If a step below does not work, check the troubleshooting guide for common fixes.
For project conventions and architecture guidance, refer to these canonical files before making changes:
- AGENTS.md — repo navigation and architecture pointers
- CLAUDE.md — working conventions and project rules
- Feature architecture standard — structure for new features
- Debugging runbook — launch and teammate diagnostics
1. Run from source or download
Download the packaged app for macOS, Windows, or Linux from the download page - no prerequisites needed. Start with the free model with no auth, or connect provider auth from the UI when you want more models.
Or run from source for development:
Requires Node.js 24.16.0 LTS and pnpm 10+. On macOS, official Node.js 24 prebuilt binaries require macOS 13.5+.
git clone https://github.com/777genius/agent-teams-ai.git
cd agent-teams-ai
pnpm install
pnpm devpnpm dev starts the desktop Electron app with hot reload. This is the default development target. Do not start a browser web dev server for normal development — the browser path lacks the full desktop IPC, terminal, provider auth, and team lifecycle behavior.
2. Open or create a project
Launch the app and select the project directory you want agents to work in. Agent Teams reads local project files and runtime/session state so the UI can show tasks, logs, diffs, and teammate activity.
TIP
Pick a Git-tracked project for the best experience. Worktree isolation and diff-based review both rely on Git.
Before launching a team, check that the project has a clean enough baseline:
git status --shortYou do not need a perfectly clean tree, but you should know which changes are yours before agents start editing. This makes task diffs and hunk-level review much easier to trust.
3. Choose a runtime path
The setup flow auto-detects installed runtimes on your machine. A common first setup is:
| Runtime | Good for |
|---|---|
| Claude | Claude Code users and existing Anthropic access |
| Codex | Codex-native workflows and OpenAI access |
| OpenCode | Free model with no auth, multi-model teams, and many provider backends |
INFO
Gemini is available as a supported provider path. See Providers and runtimes for auth options and current provider status.
See Runtime setup for detailed configuration per provider.
To verify a paid or account-backed runtime outside the app, check the binary and test auth:
# Check that the runtime is installed and on PATH
command -v claude && claude --version
command -v codex && codex --version
command -v opencode && opencode --versionIf the command fails, fix the runtime installation or PATH first. Team prompts cannot work around a missing binary or missing provider auth for models that require it.
TIP
If the binary is found but the app reports "not logged in", the environment may differ between your terminal and the app. See the auth diagnostic log to compare them.
4. Create your first team
Create a team with a lead and one or more specialists. Keep the first team small: one lead, one implementation agent, and one review-oriented agent is enough to validate the workflow.
See Create a team for the recommended structure and tips.
For the first launch, prefer a team shape like this:
| Member | Responsibility | Notes |
|---|---|---|
| Lead | Split the goal into tasks and coordinate status | Keep on the most reliable provider you have |
| Builder | Implement scoped tasks | Give clear file or feature boundaries |
| Reviewer | Review completed work | Ask it to focus on regressions and missing tests |
Avoid starting with five or more teammates. More agents increase concurrency, logs, provider usage, and conflict risk before you know the setup is healthy.
5. Give the lead a concrete goal
Write the goal like you would brief an engineering lead:
Improve the onboarding flow. Split the work into tasks, keep changes small, and ask for review before broad refactors.Good first prompts include concrete scope, safety boundaries, and verification:
Improve the docs quickstart. Keep edits inside landing/product-docs. Add practical examples, preserve existing VitePress syntax, and run `pnpm --dir landing docs:build` before marking tasks done.Avoid vague prompts such as "make the app better" for the first run. The lead can break down large goals, but better input produces smaller tasks and cleaner review.
TIP
If the team launches but no tasks appear, check whether the lead received your prompt. See agent replies are missing for diagnostics.
The lead creates tasks, assigns work, and coordinates teammates. You can watch progress on the kanban board and intervene with comments or direct messages at any time.
6. Review results
Open completed or review-ready tasks, inspect the diff, and accept, reject, or comment on individual changes. Use task logs when you need to understand why an agent made a choice.
See Code review for the full review workflow.
Before approving the first task, check three things:
- The task comment explains what changed
- The changed files match the task scope
- The verification result is visible in the task comment or logs
Common pitfalls
| Symptom | Likely cause | Check |
|---|---|---|
| App does not detect a runtime | Binary not on PATH, or app and terminal see different environments | Run command -v <runtime> in a terminal, then use the same terminal env to launch the app |
| Team launch hangs | Missing provider auth for a paid/account model, wrong model string, or runtime binary not found | See Troubleshooting |
OpenCode lane stuck on registered | Lane evidence not committed yet, or model string mismatch | Inspect ~/.claude/teams/<team>/.opencode-runtime/lanes/ |
| Agent replies missing | Runtime delivery retry, parsing, or task attribution issue | Open task logs and check the delivery ledger |
| Provider returns 429s | Rate limit reached | Wait for reset or switch model/provider |
Next steps
- Create a team — recommended team shapes and brief writing
- Runtime setup — provider auth and model selection
- Code review — review, approve, or request changes
For contributors
If you are modifying Agent Teams or these docs, start with the canonical project files at the repo root:
- CLAUDE.md — working conventions and project rules
- AGENTS.md — navigation layer for architecture and implementation guidance
- AGENT_CRITICAL_GUARDRAILS.md — hard implementation guardrails
- Feature architecture standard — structure for new features
- Agent team debugging runbook — launch, bootstrap, and teammate diagnostics
To verify this documentation site builds correctly:
pnpm --dir landing docs:build