01
FoundationDiscovery to operations
Start from a raw idea or a scoped task and move through PRD, technical spec, implementation, QA, and review without losing the thread.
Spec-driven development for AI-assisted teams
AI SDLC Kit organizes planning, implementation, QA, and review into a disciplined workflow with explicit checkpoints, reusable artifacts, and multilingual documentation.
Core Capabilities
Start from a raw idea or a scoped task and move through PRD, technical spec, implementation, QA, and review without losing the thread.
The kit treats approval gates as part of the system, not as optional ceremony, so risky assumptions are caught before code lands.
Use the built-in prompts, templates, and artifact contracts to keep outputs consistent across planning and implementation.
Pipeline
A raw idea becomes a structured, unbiased product spec.
User journeys, personas, and business rules are formalized.
Requirements gain measurable acceptance criteria and scope boundaries.
Technical spec, ordered epics, and per-epic artefacts are generated.
Implementer writes code. QA validates. Reviewer approves.
Observability, anomaly patterns, and deploy pipeline are closed.
A raw idea becomes a structured, unbiased product spec.
User journeys, personas, and business rules are formalized.
Requirements gain measurable acceptance criteria and scope boundaries.
Technical spec, ordered epics, and per-epic artefacts are generated.
Implementer writes code. QA validates. Reviewer approves.
Observability, anomaly patterns, and deploy pipeline are closed.
Agents
Structures raw ideas without technical bias. The first agent in Flow A.
Bridges product and engineering — turns the functional spec into a full PRD.
Produces the full technical architecture, epics breakdown, and per-epic artefacts.
Structures a raw task description into a scoped, unambiguous deliverable.
The kit — agents, prompts, templates, and checkpoints — is open and ready to use.