Skip to content

Product Agent

Agent deep dive · User GuideAgentsdeep dives · Technical reference: product-agent

The aidlc-product-agent is your product manager and business analyst. It transforms raw business needs, user requests, and domain knowledge into structured requirements, prioritized user stories, and well-defined scope boundaries. It ensures that every downstream artifact traces back to a validated requirement, bridging the gap between what stakeholders want and what developers build.

The aidlc-product-agent leads five stages spanning Ideation and Inception. It is the primary agent you interact with early in a workflow, asking clarifying questions about your intent, defining what is in and out of scope, and producing the requirements and stories that drive all subsequent design and implementation.

Stages Led

Stage Phase Description
1.1 Intent Capture & Framing Ideation Captures your project intent and stakeholder context
1.2 Market Research Ideation Competitive analysis and build-vs-buy assessment
1.4 Scope Definition Ideation Defines scope boundaries and prioritized intent backlog
2.3 Requirements Analysis Inception Produces structured functional and non-functional requirements
2.4 User Stories Inception Creates user stories with acceptance criteria from personas

Stages Supported

Stage Phase Contribution
1.6 Rough Mockups Ideation Validates mockups against captured intent
1.7 Approval & Handoff Ideation Validates initiative brief completeness
2.5 Refined Mockups Inception Validates mockups against user stories

What to Expect

When the aidlc-product-agent is active, expect structured questions about your project goals, target users, priorities, and constraints. It uses the tri-mode question flow (Guide Me, Edit File, or Chat) and asks targeted questions to surface ambiguities and fill gaps. It prioritizes ruthlessly — helping you distinguish must-have from nice-to-have.

How It Collaborates

The aidlc-product-agent works closely with the aidlc-architect-agent on feasibility and dependencies, the aidlc-design-agent on UX alignment, and the aidlc-delivery-agent on capacity and scope validation. Its outputs (requirements, stories, scope) are consumed by nearly every downstream agent.

Key Principles

  • Every requirement must trace to a stakeholder need — no invented requirements
  • If a requirement cannot be verified through a test, it is not a requirement
  • Ambiguity is the enemy — when something seems obvious, confirm it
  • Value over volume — fewer well-defined stories beat a large vague backlog
  • Stories should cut vertically through all layers, not horizontally