onboard-product-owner

name	onboard-product-owner
description	Use when configuring the Product Owner agent for a specific project. Gathers issue tracker details, backlog norms, and current roadmap state, then writes them to the Product Owner's project memory. Run /onboard first to establish shared project context. This is the reference implementation for per-agent onboarding skills.
user-invokable	true
argument-hint
allowed-tools	Read, Glob, Write
context	fork

Onboard — Product Owner

Configure the Product Owner agent for your project by capturing issue tracker details, backlog norms, and current roadmap state.

Prerequisites

This skill layers on top of shared project context. Before starting, read:

.claude/agent-memory/engineering-leaders/PROJECT.md

Track whether this file exists as shared_context_exists. If it does not exist, note:

"Shared project context hasn't been set up yet. I can continue with Product Owner-specific questions, but running /onboard first will give all agents a better foundation. Would you like to continue anyway?"

• If the user says yes: track shared_context_exists = false and continue to the Output Location section.
• If the user says no: stop here. Do not proceed to the interview or write any files. Tell the user: "No problem. Run /onboard first to set up shared project context, then come back to /onboard-product-owner."

Output Location

Product Owner project memory is written to:

.claude/agent-memory/engineering-leaders-product-owner/MEMORY.md

If this file already exists, show the user a brief summary of what is recorded (issue tracker platform, current phase, sizing convention) and ask:

"Product Owner memory already exists. Would you like to:

(a) Update specific sections — I'll ask which sections to replace and re-run only those questions. All other sections are preserved unchanged. (b) Start fresh — I'll run the full interview and replace the entire file when complete. If you abandon the interview before finishing, the original file is left unchanged."

If the user chooses (a):

Show the list of sections (Issue Tracker, Roadmap and Phases, Team Norms) and ask which to update. Re-ask only the questions for those sections (Q1-Q3 for Issue Tracker, Q4-Q6 for Roadmap and Phases, Q7-Q9 for Team Norms), then merge the new answers into the existing file by replacing only those sections. Sections not selected are preserved verbatim from the original. Any content in the file that does not correspond to a template section (e.g., manually added sections) must also be preserved verbatim — do not discard unrecognized content.

If the user chooses (b):

Proceed through the full interview below. Write the file only after all questions are answered. The interview is considered abandoned if the user explicitly says to stop (e.g., "stop", "cancel", "never mind", "let's come back to this") or leaves the conversation without completing the Write section. If abandoned, do not write anything and tell the user: "Interview not completed. The original file is unchanged."

Process

Introduce yourself before asking anything:

"I'll ask you a few questions about how your team tracks and manages work — one at a time. This gives the Product Owner agent the context it needs to advise on sequencing, write well-structured stories, and track your roadmap. Skip any question with 'skip' or 'not sure yet'."

Issue Tracker

Q1 — Issue tracker platform

"Which issue tracker does your team use?

(a) GitHub Issues (b) Jira (c) Linear (d) Notion (e) Other — I'll describe it (f) None / we don't use one"

Q2 — Project or board URL

"What's the URL for your project board or repository? (e.g., https://github.com/org/repo or https://linear.app/team/project)"

Skip this question if the user answered (f) above.

Q3 — Issue structure

"How are work items organized in your tracker? For example: do you use epics, labels, milestones, sprints, or story points? Describe what you actually use."

Roadmap and Phases

Q4 — Current phase or milestone

"What is the team focused on right now? Is there a named phase, milestone, or goal you're working toward?"

Q5 — Immediate priorities

"What are the top 2–3 things the team is trying to ship or complete in the near term?"

Q6 — Known blockers or risks

"Are there any known blockers, open decisions, or risks that could affect sequencing? (Skip if none)"

Team Norms

Q7 — Story sizing convention

"How does your team size or estimate work items?

(a) T-shirt sizes (XS/S/M/L/XL) (b) Story points (Fibonacci or similar) (c) Days / hours (d) Just small/medium/large (e) We don't size work items"

Q8 — Definition of Done

"What does 'done' mean on your team? Are there specific criteria a story must meet before it's closed? (e.g., code reviewed, tests passing, deployed to staging, acceptance tested)"

Q9 — Anything else

"Is there anything else the Product Owner should know about how your team works — norms, constraints, or preferences that aren't captured above? (Skip if nothing comes to mind)"

Write Product Owner Memory

After collecting all answers, write the memory file. Use the structure below, omitting sections where the user skipped or had nothing to say.

# Product Owner — Project Memory

> Configured by /onboard-product-owner. Update by re-running the skill or
> editing this file directly.

## Issue Tracker

- **Platform:** [from Q1]
- **Project URL:** [from Q2, if provided]
- **Issue Structure:** [from Q3]

## Current Roadmap State

- **Current Phase / Milestone:** [from Q4]
- **Immediate Priorities:** [from Q5 — bullet list]
- **Known Blockers / Risks:** [from Q6, if provided]

## Team Norms

- **Story Sizing:** [from Q7]
- **Definition of Done:** [from Q8]
- **Other Norms:** [from Q9, if provided]

If shared_context_exists = true, append this section to the file:

## Shared Project Context

See `.claude/agent-memory/engineering-leaders/PROJECT.md` for project
overview, tech stack, team structure, and SDLC process.

If shared_context_exists = false, append this section instead:

## Shared Project Context

Not yet configured. Run `/onboard` to set up shared project context (project
overview, tech stack, team structure, SDLC process) that all agents can use.

Create the directory .claude/agent-memory/engineering-leaders-product-owner/ if it does not exist, then write the file.

Confirmation

Present a confirmation summary:

Product Owner configured.

Memory written to:
  .claude/agent-memory/engineering-leaders-product-owner/MEMORY.md

The Product Owner now knows:
  - Issue tracker: [platform from Q1] at [url from Q2, or "not specified"]
  - Current focus: [phase/milestone from Q4]
  - Top priorities: [first item from Q5, or "see memory"]
  - Sizing: [convention from Q7]
  - DoD: [one-line summary from Q8, or "not specified"]

The Product Owner agent is ready. Invoke it by saying "consult the product
owner" or asking about roadmap, sequencing, or story authoring.