aibdd-form-entity-spec

You can generate DBML from feature specs by running a reconciler-based approach that derives the desired erm.dbml data model state from your Feature Files for both greenfield and existing projects.

A reconciler computes data model diffs by deriving the desired erm.dbml state from Feature Files, reading the current erm.dbml, and calculating the differences to output a change plan with IMPL_IMPACT.

Yes, you can infer Aggregate, Enum, and Relationship structures from Feature Files into DBML format automatically, with conflict reporting to trigger coordination when needed.

The best way to update an existing data model is using a reconciler contract that reads your current erm.dbml, derives the desired state from feature specs, and computes diffs for a structured change plan.

Yes, you need to provide the feature specs path to support /discovery and carry-on invocations, ensuring the caller supplies necessary paths and execution plans to generate erm.dbml.

DBML generation requires a reconciler contract to enforce compliance when deriving the data model state, reading references like reconciler-contract.md, and computing diffs for the output change plan.