OpenRig Software Factory

OpenRig Software Factory

Start with a useful repository outcome and add coordination when it earns its cost. A beginner can complete reviewed work without Workflow. These are choices using existing capabilities, not stages everyone must graduate through.

Choose how the team works

For the concrete queue loop, wake behavior and optional two-slice Workflow, read references/worked-example.md. Installed copy: rig context get skills/core/openrig-software-factory/references/worked-example.md. A roadmap, YAML file or wake does not execute work or authorize a new outcome.

Choose the first team and its providers

Ask what the user wants to build and which working account(s) they have: Claude Code, Codex, or both. Recommend one of three teams: starter (a Claude builder and a Codex reviewer, for one bounded change), workshop (a lead, a builder, QA and a reviewer; a rig bundle installed from its pinned listing) or factory (seven agents for sustained product work). When the user lacks a provider a team needs, write an adapted copy of the team under the same name, as the kernel operator's guidance describes; never offer per-provider variants. first-project is starter's old name. Check only the CLIs/logins the team needs; request claude auth login or codex login once when that selected login is missing. No credential copying, unused provider prerequisite or silent model/provider fallback.

Read the compatible getting-started guide's Choose your providers and Start the kernel and check its state sections before launch. The choice selects the two project agents. Kernel auto-boot independently uses available authenticated accounts, so it may use both even when the project uses one. Do not add an unused-provider login gate or manual kernel setup to this path. An instance-wide provider restriction is a separate request. Preserve an existing kernel and working user rigs.

Show the chosen team, resolved runtimes/models and exact rig up <team> --cwd . --plan / rig up <team> --cwd . commands. Codex seats retain gpt-6-astra; Claude seats use the configured native default without an OpenRig model override. Confirm that model with the user and its availability; verify the native session's actual model before consequential work. Use the chosen rig name in owner/checker addresses (in the starter, dev-build@<rig> and dev-review@<rig>) throughout the same task and return.

Establish the working agreement

Read the repository instructions, current work and desired user-visible result. Verify the intended instance, code/work roots, real seat addresses and native readiness. Reuse a suitable small team; an existing agent can bootstrap it. A kernel operator is optional and is not automatically the project owner.

Agree the work boundary, time/spend limit, who answers unresolved choices, and when to stop: checked result, no authorized next work, exhausted budget, or a real user/permission/provider blocker. Background daemon checks are not themselves model turns, but delivered wakes and resumed work can spend tokens. Prefer an event-driven wait to frequent empty reminders. Wakes cannot answer a user question, clear a permission prompt or guarantee progress.

Ask once before launching or assigning work. For a team with no permission policy, seat choice or named Codex profile, recommend keeping the team default: Claude team seats run ordinary rig commands, project reads and common tests without prompts, and lifecycle commands such as rig up and rig down still ask. Only if they want more, offer: “Remember these selected OpenRig commands in your native settings for this project?” Yes / No — keep the team default. Reuse an existing explicit choice for this scope. Explain that a remembered allowance can cover all rig verbs, but Claude team seats still ask before lifecycle commands, at personal project scope unless the user explicitly chooses user-wide sessions. It is not global YOLO or permission to invent work. On an actual Yes, follow Applying a permission policy to add existing native rules, preserve stricter/unrelated settings, and verify the target conversation. No or no answer leaves settings alone and keeps the team default. Remember the explicit choice and exact additions in the existing onboarding context; “Undo the OpenRig command allowances added by this setup” removes only those additions. Broader access remains a separate opt-in.

Keep purpose, acceptance, decisions and evidence in existing project files. Deliver selected context and obtain each seat's scope reaction; retrieval alone is not peer delivery. The owner carries the candidate through the chosen check and bounded repairs, reports how to try it, and retains the next authorized task or explicitly reports none. Preserve work and custody before a supported stop.