speckit-plan

To generate implementation plans from a feature specification, you can automate the process by converting technical context and requirements into design artifacts like a data model and contracts. This resolves project unknowns and enforces gate rules before building.

The best way to resolve unknowns in technical context is to detect items needing clarification and generate research tasks to address them. This ensures all technical context is fully clarified before proceeding to create the data model and contracts.

Data models and contracts are generated from a feature spec by running a structured Phase 0 to 1 workflow. This process turns technical context and requirements into concrete files including research.md, data-model.md, and a contracts directory when external interfaces exist.

Yes, you can use extension hooks to customize the planning pipeline by running before_plan and after_plan hooks. This extends the workflow without modifying the core process, provided the hooks are present and enabled in your environment.

Implementation planning fails on unresolved issues because the workflow enforces gate rules by erroring on unresolved or unjustified issues. This prevents moving forward until all technical context unknowns are properly clarified and justified.

Yes, spec-kit planning supports updating agent context after generating design artifacts. The workflow executes an update-agent-context script alongside producing quickstart.md and other files to ensure the agent has the latest project context.