Skip to content

create-issue.md

Latest commit

 

History

History
180 lines (99 loc) · 3.63 KB

create-issue.md

File metadata and controls

180 lines (99 loc) · 3.63 KB

Command: /create-issue

Required Knowledge

Load only these knowledge files before executing:

  • knowledge/product-principles.md
  • knowledge/product-lessons.md

Purpose: Capture a raw product idea and convert it into a structured issue that can enter the product development pipeline.

Role

You are responsible for converting messy ideas into structured problem statements.

You must think like a product manager discovering opportunities.

Input

You will receive a raw idea.

Examples:

"AI PM portfolio generator"

"Tool that summarizes Gmail and sends WhatsApp notifications"

Process

Follow this structure.

0 Pre-Flight Questions

Before generating the issue, assess whether the raw idea is thin (missing user, behavior, or outcome).

If thin: Ask 2–3 targeted questions in a single message. Wait for answers before proceeding.

Questions to ask when needed:

  • What's the current behavior / pain point?
  • What does the desired outcome look like?
  • Who is the primary user?

If fully formed: Skip questions and generate directly.

Keep questions brief — one message, max 3 questions, no back-and-forth.

Issue Type (always required): Infer from the raw idea if obvious; otherwise include as one of the clarifying questions.

  • Feature — new capability that didn't exist before
  • Enhancement — improvement to an existing capability
  • Bug Fix — something broken that needs fixing

Store the result as issue_type in the issue file header.

1 Problem Statement

Describe the problem clearly.

Current State: What is happening today / what pain exists.

Desired Outcome: What should be true after this is solved.

2 Target User

Identify the primary user.

3 Why This Problem Matters

Explain why solving this problem creates value.

4 Opportunity

Describe the opportunity if the problem is solved.

5 Initial Hypothesis

Define a simple hypothesis.

Example:

"If we build X for Y users, it will solve Z problem."

6 Risks / Open Questions

Call out anything that could block this or make it hard to build.

  • Dependencies on external systems, APIs, or other issues
  • Regulatory, technical, or product unknowns
  • Questions that /explore should resolve

Omit this section if nothing notable applies.

Output Format

Return the result in this structure.

Issue Title Type: Feature | Enhancement | Bug Fix

Problem (Current State → Desired Outcome)

User

Why it Matters

Opportunity

Hypothesis

Risks / Open Questions (omit if none)

Post-Output Steps

After writing experiments/ideas/issue-<NNN>.md and updating project-state.md:

Auto-Write CHANGELOG Entry

Append to the top of CHANGELOG.md immediately after writing the issue file:

## YYYY-MM-DD — Issue Created: issue-NNN
- **Type**: Feature | Enhancement | Bug Fix
- **Title**: <issue title>
- **App**: <project/app name if known, else TBD>
- **Status**: Discovery

Use today's date. Do not modify any other CHANGELOG content.

Auto-Bind Linear

Immediately run /linear-bind for the new issue.

This ensures linear_enabled: true is set from the moment every issue is created — no manual bind step required.

Expected outcome:

  • experiments/linear-sync/issue-<NNN>.json created with team, project, and root issue ids
  • project-state.md updated with linear_enabled: true and all linear binding fields
  • Linear root issue created in Backlog with label AI Product OS/Discovery

If Linear is unavailable, log the failure explicitly and continue — do not block the pipeline.

Next Step

Send this issue to the Research Agent for validation via /explore.