[M3 Request] processing image inputs (like screenshots)

[braindamagedjustnow]

Capability area

Agent harness / Agent Teams

What does M2.7 fail to do for you?

Using it as local model with Cursor I lack screenshot porcessing

What would "good" look like in M3?

Ability to process image inputs like screenshots and others

References

No response