Install ProPR, or skip the server
Run npm install -g propr-cli and propr setup on a Linux Docker host (quickstart). ProPR Hosted runs a dedicated instance for you when you would rather not run one.
ProPR: the open-source alternative to Factory
Factory builds one agent, Droid, and sells it as a hosted platform with desktop, CLI, web, and chat surfaces. ProPR is the Apache 2.0 delivery path for agents you already subscribe to, on your own server.
vs
Factory.ai
Who supplies the agent
Factory supplies the agent and the platform, with private VPC and air-gapped deployment reserved for its Enterprise tier. ProPR supplies the process: planning, isolated runs, PR delivery, and review loops, self-hosted in every installation.
Factory pathSlack, Linear, or Jira → Droid in Spec Mode → pull requestFollow-ups go back through the thread that started the work.
ProPR pathPlan → isolated workspace → GitHub PR → follow-up commitPlans become issues, results become PRs, review comments become commits.
Switch from Factory to ProPR.
What you keep: your GitHub repositories and review process, and the Claude or ChatGPT plan ProPR runs through Claude Code or Codex; Factory takes API keys only.
The same ideas, ProPR's names
| Factory | ProPR |
|---|---|
| Droid | Any of five agents, chosen per task |
| Spec Mode | Plan, approved before code runs |
| Delegation from Slack, Linear, or Jira | GitHub issue, web UI, CLI, or MCP |
@droid review | /review and /fix |
| Follow-up through the originating thread | Any plain PR comment |
| Factory's model router | Agent and model label per task |
From zero to the first pull request
Add the same repository and sign in your agent
Open the web UI, sign in with GitHub, and add the repository. Sign in Claude Code, Codex, or Antigravity CLI with the plan you already hold.
Start the first task and approve the plan
Describe the feature in Planner Studio, approve the plan, and create its issues; or add the AI label to one issue.
Review the pull request in GitHub
Review the diff the way you review any pull request. A plain comment on it becomes the next commit on the same branch; /review posts scored findings and /fix applies them. The epic playbook lands a whole plan as reviewed pull requests.
What stays with Factory: ProPR has no Slack, Linear, or Jira entry point and no vendor-managed compute; it runs on your Linux Docker host or a ProPR Hosted plan.
Factory popularized the "software factory" framing: an agent-native platform where Droid takes work from ticket to pull request across the desktop app, CLI, web, Slack, Linear, and Jira. Spec Mode has Droid write a plan before a multi-step implementation, and you approve it with an autonomy level. The Factory Droid GitHub App authors the pull request, and a GitHub Actions review workflow posts inline comments on PR events or on an @droid review mention. Follow-up instructions go back through the Slack, Linear, or Jira thread that started the work. You pick the model from Anthropic, OpenAI, Google, xAI, and open-weight options, or bring your own API keys through custom models. Individual tiers run $20–$200 per month; Teams starts at $60 plus $40 per seat, and Enterprise adds SSO, audit controls, dedicated compute, and private or air-gapped deployment.
ProPR agrees with the premise and changes who owns it. The loop runs as free, open-source software on your own server: planning, parallel isolated runs, scored review, PR delivery, follow-up commits. The agents are the ones you already pay for: Claude Code on a Claude plan, Codex on a ChatGPT plan, or any of the five on your own API keys, with no token markup. A human merges, or you switch on auto-merge for green pull requests. The software factory page maps the full loop.
Factory fits best when...
- You want a vendor-built agent with a managed platform, model router, and compute around it.
- Engineering work starts in Slack, Linear, or Jira, and those threads drive the agent.
- You buy through an Enterprise contract with SSO, audit controls, and vendor-managed, private VPC, or air-gapped deployment.
ProPR fits best when...
- You run the agents you already subscribe to, switch them per task, and compare them on the same issue.
- Repositories, logs, credentials, and execution live on infrastructure you run from day one, for free.
- GitHub drives delivery: plans become issues, results become PRs, and plain review comments become follow-up commits.
Hosted Droids versus a self-hosted delivery path.
| Question | Factory.ai | ProPR |
|---|---|---|
| Open source | No | Apache 2.0 |
| Self-hosted on the free tier | No; private VPC and air-gapped on Enterprise | Yes |
| Ships its own proprietary agent | Yes, Droid | No, it routes your agents |
| Plan before implementation | Spec Mode | Reviewed plan on every task |
| Runs on AI subscriptions you already hold | API keys via custom models; no subscriptions | Yes |
| Work ships as GitHub pull requests | Yes | Yes |
| Review comments become follow-up commits on the same branch | Review comments only; follow-ups from the Slack, Linear, or Jira thread | Any plain comment, any agent |
| Platform pricing beyond model costs | $20–$200/mo tiers; Teams $60 plus $40 per seat | Free software |
Verified November 2026: factory.com/pricing, Spec Mode, code review automation, delegations, models, network and deployment.