A separate place for still-image edits in Civon
Published by: ethanjamescolez, Founder of Civon
An existing product boundary in Civon is worth making explicit: editing a still image has its own browser route. This note is prepared for the product owner with AI-generated writing. It describes the current workflow, rather than announcing a new release or a growth milestone.
The image editor starts with supported reference images and a written request for a change. The user selects an available model and compatible output settings, checks the credit quote, then generates, inspects and downloads a still-image result. The number of usable references and the available settings depend on the selected model and workflow.
That leaves an important gap between producing a result and accepting it. For a hypothetical product photograph, a warmer wall might be the intended change, while the label, outline and material details should remain recognizable. A successful generation does not establish that those details survived. Faces, text and geometry can change unexpectedly, and the result needs comparison with the source.
For makers designing generated-media interfaces, what should make acceptance more deliberate: keeping the source visible beside the candidate, asking the user to name the details to preserve, or providing a short manual checklist? These are design questions, not claims that those controls have shipped in Civon.
Inputs still need the relevant rights and permissions. Current availability, settings, quotes and saving conditions should be checked before use; there is no promise of exact edits or cleared commercial rights.