Clarification: FastCRUD methods assume primary key as “id” — not working for models with custom primary key fields

[ujjawal-ibm]

Description

Hi team 👋,

I wanted to clarify and possibly request enhancement support for FastCRUD when using models that have a primary key field name other than id.

Currently, the default methods like fast_crud.get(db, id=value) or fast_crud.update(db, id=value, obj=value) internally assume that the primary key column is named id.
If our SQLModel has a custom primary key (for example, fsid), these methods fail to locate the row.

Example

from fastcrud import FastCRUD
from sqlmodel import SQLModel, Field

class Cluster(SQLModel, table=True):
    fsid: str = Field(primary_key=True)
    name: str

ClusterCRUD = FastCRUD(Cluster)

# This call fails
ClusterCRUD.get(db, id="abcd-1234")

It gives the error:
KeyError: 'id'

Expected Behavior

Ideally, FastCRUD should detect and use the correct primary key field (in this case, fsid) automatically, similar to how SQLModel identifies it internally.

Question

1. Is this an intentional design limitation or a current missing feature?
2. If not supported yet, can we expect a planned fix or enhancement in upcoming versions?

Workaround Tried

For now, I’m using explicit SQLModel queries:

result = await db.execute(select(Cluster).where(Cluster.fsid == fsid))

I’d be happy to test or contribute a small PR if guidance can be provided on the preferred direction for supporting custom primary key fields.

[igorbenav]

@ujjawal-ibm try running

ClusterCRUD.get(db, fsid="abcd-1234")

FastCRUD uses kwargs to define filters, so you shouldn't have any problem just passing what your field is actually named. Essentially you may just pass as keyword arguments no matter what the filters are named