Skip to content

First Ten Minutes

This guide gets you from zero to your first implemented task in about ten minutes. Follow the path that matches your situation.

Terminal window
aios init my-project
cd my-project
Terminal window
aios next

aios next is read-only. It prints a recommendation based on what files exist.

Step 3: Ask your agent to discover your product

Section titled “Step 3: Ask your agent to discover your product”

Paste this into your AI coding agent:

Read docs/context/context-map.md, then use the product-discovery skill
or .aios/prompts/00-discover-product.md to interview me about my product idea.
Fill docs/product/vision.md with the result.

Pause and review. Read docs/product/vision.md. Make sure it captures your idea accurately before continuing.

Read docs/product/vision.md and use the prd-generator skill
or .aios/prompts/01-generate-prd.md to create a PRD.
Save it as docs/product/prd.md using .aios/templates/prd.template.md.

Pause and review. Check that acceptance criteria are testable and that scope is realistic.

Terminal window
cd your-existing-project
aios adopt
aios validate

aios adopt adds structure without overwriting existing files.

Terminal window
aios next

Step 3: Start with product discovery or jump in

Section titled “Step 3: Start with product discovery or jump in”

If you have a clear idea, skip to Continue the Lifecycle. Otherwise, ask your agent:

Read docs/context/context-map.md, then use the product-discovery skill
or .aios/prompts/00-discover-product.md to interview me about what we are building.
Fill docs/product/vision.md with the result.

Pause and review. Verify the vision before moving on.

After vision and PRD are ready, follow these prompts in order. Pause after each one to review the output.

Use the architecture-design skill or .aios/prompts/02-generate-architecture.md
to create architecture context. Save it using .aios/templates/architecture.template.md.
Use the adr-generator skill or .aios/prompts/03-generate-adr.md to record
this decision: <your decision here>. Save it using .aios/templates/adr.template.md.

Pause and review. Make sure architecture and decisions align with your constraints.

Use the task-breakdown skill or .aios/prompts/04-generate-tasks.md to split
the PRD into small implementation tasks. Save them in docs/tasks/.
Use the implementation-planner skill or .aios/prompts/05-plan-implementation.md
to create a plan for the first task.
Use the task-implementation skill or .aios/prompts/06-implement-task.md
to implement the first task. Run npm run build or your project tests
to verify.

Pause and review. Check the diff. Make sure the change is scoped and correct.

Use the code-review skill or .aios/prompts/07-review-code.md to review
the changes against the task acceptance criteria.

UI/UX Design (when building user-facing features)

Section titled “UI/UX Design (when building user-facing features)”
Use the ui-ux-design skill or .aios/prompts/13-design-ui-ux.md to design
user flows and interface states for <your feature>.
Save it using .aios/templates/design.template.md.

If something fails during implementation or testing:

Read the error output and the task acceptance criteria.
Identify what failed, fix the issue, and re-run the test or build.
Update docs/reviews/ if the fix changes the implementation approach.

The full loop is:

  1. Discover product
  2. Generate PRD
  3. Design architecture
  4. Record decisions (ADRs)
  5. Break into tasks
  6. Plan implementation
  7. Implement one task
  8. Review the change
  9. Repeat from step 6

Each step produces a file. Each file becomes context for the next step. The agent reads the right document instead of relying on chat memory.

You should pause and review after these steps:

  • Product vision: before the agent writes a PRD based on wrong assumptions
  • PRD: before architecture is designed around the wrong scope
  • Architecture/ADR: before tasks are created with incorrect technical direction
  • Task implementation: before moving to the next task with an unverified change

Do not skip reviews. The files are meant to be read by humans, not just by agents.