Guide · Agent skill
How to use the Development Team Pod Workflow agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The development-team skill is a set of written instructions that an AI agent loads when it is working as part of a development pod: a group of agent seats with builder, QA and design roles inside the OpenRig framework. It tells the agent how to start from an assigned task, read the project's planning files, build and verify a change, and hand the result off. This guide explains what the skill does, how to install it, how to trigger it, and what it needs, so you can decide whether it fits your setup.
Development Team Pod Workflowmvschwarz/openrig · Apache-2.0 Download the skill Quick start
- 1Download and unzip. Download the skill ZIP from this page and unzip it so the folder contains SKILL.md and references/refactoring.md.
- 2Place the folder. Move the unzipped skill into ~/.claude/skills/ for personal use or into the project's .claude/skills/ folder, or upload it under Settings → Capabilities → Skills in the Claude apps.
- 3Check the environment. Confirm the rig command is available by running rig whoami --json in your OpenRig environment.
- 4Confirm planning files. Make sure project.yaml, mission.yaml and slice.yaml exist for the work you want the pod to do.
- 5Make a request. Describe the task in plain terms, naming the role, such as implementation, QA handoff or design.
- 6Review the handoff. Check that the agent reports the candidate, commands, observed results and remaining uncertainty.
What is an agent skill, and what does this one do?
An agent skill is a folder containing a SKILL.md file. The file holds instructions the agent reads when a request matches the skill's description. This skill's description reads: use when a development pod begins or hands off implementation, QA, or design work.
Once loaded, the skill shapes how the agent works. It asks the agent to start from the assigned task, resolve the project's planning files, reproduce a bug before fixing it, test through public interfaces, and report exact evidence and remaining uncertainty instead of claiming success. For consolidation or refactoring work, it points the agent to a separate refactoring guide.
- Start from the work: run rig whoami --json, then follow project.yaml to mission.yaml to the active slice.yaml to the selected component or wave map.
- Build and verify: reproduce defects first, verify through the public surface, and report the candidate, commands and observed results.
- QA and design: QA compares delivered behavior with the actual contract; design clarifies user flows when the outcome needs it.
- Blockers and return: use systematic-debugging for unexpected failures and verification-before-completion before claiming success; send the candidate and evidence through the selected handoff.
Install the development-team skill
Download the skill ZIP from this page and unzip it. The skill folder contains SKILL.md and references/refactoring.md, and the folder must be placed where your agent looks for skills. You have two options.
For a personal setup, unzip the skill into ~/.claude/skills/. For a single project, unzip it into the project's .claude/skills/ folder. Alternatively, in the Claude apps you can upload the skill under Settings → Capabilities → Skills.
- Personal: ~/.claude/skills/ (available across your projects)
- Per project: .claude/skills/ inside the project directory
- Claude apps: Settings → Capabilities → Skills, then upload
Prepare the OpenRig environment and planning files
The skill assumes an OpenRig environment in which the rig command is available. The lookup flow also needs three planning files: project.yaml, mission.yaml and slice.yaml. Without them, the agent has no path to resolve, and most of the skill will not apply.
The skill points to the full lookup and precedence rule in docs/reference/product-journey-sdlc.md under the heading resolve the selected path. If OpenRig is installed, the same reference is at $OPENRIG_HOME/reference/product-journey-sdlc.md.
- OpenRig environment with the rig command available
- Planning files: project.yaml, mission.yaml, slice.yaml
rig whoami --jsonTrigger the skill with natural requests
You do not need a special command. The skill loads when the request matches its description, so phrase requests around the pod's work: starting an implementation task, handing off to QA, or beginning design work. Naming the role and the task makes it more likely to match.
If the agent does not appear to follow the workflow, say so directly and point it at the skill, for example by asking it to apply the development-team skill to the current task.
- Implementation: "Start the builder work for slice 3 using the development team workflow."
- QA handoff: "Hand off this build to QA. Send the candidate, commands and observed results."
- Design: "Clarify the user flow for the new export screen before building it."
- Refactoring: "Consolidate these duplicated parsers, following the development-team refactoring guide."
Safety notes: scripts and network
The skill runs no scripts and uses no network access. It is guidance text. It does not run code or change files by itself. The commands it names, such as rig whoami --json, are run only if the agent chooses to run them in its environment, so review what the agent executes in a pod before you approve it.
A permission prompt is treated as a concrete blocker. The skill asks the agent to report the command and its consequence rather than treat the prompt as progress.
- Runs scripts: no
- Uses network: no
- Commands named in the skill run only if the agent chooses to run them
Tips for getting good results
The skill works best when the planning files already describe the outcome. Vague slices give the agent less to resolve. Ask for evidence explicitly, because the skill expects the agent to report observed results and remaining uncertainty.
Use the refactoring guide only for consolidation or refactoring work. The skill asks for it in that case, and it asks the agent to trace callers before consolidating.
- Keep slice.yaml specific about the user outcome and file territory.
- Ask for the exact commands and results in the handoff.
- Expect QA to be held by the builder on tiny changes unless independent QA was selected.
What people use it for
- A builder seat starts a slice by resolving the planning files and reproducing a reported bug before fixing it.
- A QA seat reviews a builder's diff and exercises the result against the actual contract.
- A design seat clarifies an ambiguous user flow before the builder starts work.
- A team consolidates duplicated logic and uses the refactoring guide to trace callers first.
- A maintainer reads the skill to see how OpenRig structures agent roles and handoffs.
Frequently asked questions
Is the development-team skill free?
The skill is published under the Apache-2.0 licence, as listed on this page. Check the repository's licence file for the full terms before reuse.
What do I need to use the development-team skill?
You need an OpenRig environment with the rig command available, and project planning files (project.yaml, mission.yaml, slice.yaml) for the lookup flow. The skill runs no scripts and uses no network.
Where do I install the development-team skill?
Unzip it into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder. In the Claude apps, you can also upload it under Settings → Capabilities → Skills.
Can I use this skill outside OpenRig?
Not really. It relies on OpenRig commands such as rig whoami and on files like project.yaml, mission.yaml and slice.yaml, which a general agent will not have. Much of the guidance will not apply outside OpenRig.
Does the skill change my files or run code by itself?
No. It is guidance text. Commands it names are run only if the agent chooses to run them in its environment.
Why won't my agent follow the workflow?
Check that the skill folder is in the right location and that the request matches the skill's description. Make sure the rig command and the planning files exist, since the lookup flow depends on them.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you run multi-agent development pods on OpenRig and want agents to follow a consistent build, test and handoff process. Pick something else if you work alone outside OpenRig, because the commands and planning files it relies on will be missing and most of it will not apply.
Download the skill