Built as working context, not shelfware.
This resource is meant to be useful inside the tools where product work now happens: your codebase, your notes, and your AI-assisted workflow.
01
Paste the markdown into Claude, ChatGPT, Cursor, Codex, Gemini, or another AI agent as reusable project context.
02
Use it before a planning, implementation, review, or audit session so the agent has constraints, criteria, and working structure up front.
03
Adapt the sections to your product, team, or repo before asking the agent to execute against it.
Markdown previewai-product-sprint-checklist.md
# AI Product Sprint Checklist Use this when a small team needs to move from product ambiguity to a verified shipped slice with AI support. ## 1. Frame The Sprint - [ ] Write the user problem in one sentence. - [ ] Identify the user, buyer, or operator affected by the work. - [ ] Name the current workaround or pain. - [ ] Define the desired behavioral outcome. - [ ] Define the smallest useful shipped version. - [ ] List the non-goals for this sprint. - [ ] Decide which decisions are human-owned and which tasks AI can accelerate. ## 2. Gather Context - [ ] Collect product docs, tickets, analytics, customer notes, and screenshots. - [ ] Ask AI to summarize the evidence, but verify every claim against the source. - [ ] Ask AI for missing questions, edge cases, and assumptions. - [ ] Identify the current system constraints: data, permissions, API limits, design system, dependencies. - [ ] Capture examples from competitors or adjacent products. - [ ] Create a short brief with problem, audience, outcome, scope, and constraints. ## 3. Explore Solutions - [ ] Generate 3-5 possible approaches. - [ ] Score each approach by user value, implementation cost, risk, reversibility, and learning value. - [ ] Select the most useful thin slice. - [ ] Write the core user journey. - [ ] Identify required states: empty, loading, success, error, partial, permission denied. - [ ] Decide what can be mocked, stubbed, or hardcoded for the first pass. ## 4. Design The Experience - [ ] Draft the screen hierarchy before visual polish. - [ ] Check that the primary action is obvious. - [ ] Write interface copy in plain user language. - [ ] Define responsive behavior for mobile, tablet, and desktop. - [ ] Confirm keyboard and screen-reader basics. - [ ] Ask AI to critique the design for ambiguity, hidden assumptions, and missing states. - [ ] Apply human judgment before accepting critique. ## 5. Build With AI Assistance - [ ] Ask AI to inspect the existing code patterns before proposing code. - [ ] Keep the first implementation close to existing architecture. - [ ] Make small commits or checkpoints. - [ ] Review generated code for data leaks, auth mistakes, brittle state, and unnecessary abstractions. - [ ] Add focused tests around the risky behavior. - [ ] Keep a short note of prompts or agent decisions that changed the implementation. ## 6. Verify The Slice - [ ] Run build, lint, and relevant tests. - [ ] Exercise the happy path manually. - [ ] Exercise error and empty states manually. - [ ] Check mobile and desktop layouts. - [ ] Check performance basics: large assets, blocking work, repeated network calls. - [ ] Ask AI for a final bug-risk review, then verify the findings yourself. - [ ] Document known limitations and next steps. ## 7. Ship And Learn - [ ] Confirm analytics events or success measures. - [ ] Prepare release notes or internal handoff notes. - [ ] Watch the first real usage or QA session. - [ ] Collect user feedback. - [ ] Decide whether to iterate, expand, pause, or remove.