Every phase readswhat the last one wrote.

Intake

Turn a description into a brief the pipeline can work from.

• A structured project brief from your description
• Answers to the questions your description left open
• Attached documents (PDF, DOCX, Markdown, plain text) read into context
• Background research on the domain, where it helps

Runs in the new-project wizard, before the workspace exists.

Discovery

Extract capabilities and requirements.

• Core features with priority
• User roles and the permissions each one needs
• A recommended tech stack, with the reasoning
• The entities in the domain and how they relate
• Non-functional requirements — performance, security, compliance

Specification

Define APIs, schema, and routes.

• REST endpoints with methods, paths, and auth requirements
• UI pages with routes, components, and layouts
• An authentication strategy
• Error handling patterns

Architecture

Decompose into typed components and flows.

• Typed components, each traced back to a discovery feature
• The flows between them
• An architecture diagram you can read and export
• Review findings against the specification

Database

Design the complete data schema.

• A logical data model — tables, columns, keys, indexes
• DDL emitted deterministically from that model
• Relationships and constraints

Skipped for projects with no persistence, decided by rule rather than by the model.

Planning

Break down into epics and stories.

• Epics grouping related work
• Stories with descriptions and complexity
• Acceptance criteria for each story
• The files each story touches
• Dependencies, and the order to build in

Estimation

Story points, hours, and risk assessment.

• Points and an hour range per story
• Risks, with mitigations
• A rolled-up estimate for the whole plan

Sprint plan

Organise sprints and milestones.

• Stories assigned to sprints, respecting their dependencies
• Milestones across the plan
• A delivery sequence your team can commit to