Problem

When editing existing Google Slides presentations (e.g. regenerating individual slides after content changes), the current workflow requires:

1. list-slides to get IDs — but IDs tell you nothing about content
2. Export entire deck as PDF to visually identify which slide is which
3. delete-slide + add-slide --before to replace a single slide
4. Export again to verify

This is error-prone. Without being able to read individual slide content, it's easy to delete the wrong slide (e.g. deleting Exploration Mode when you meant to delete a duplicate What Is Claude Code).

Proposed Commands

gog slides read-slide <presentationId> <slideId>

Show slide content: speaker notes text, any text elements on the slide, and image references. Even just the speaker notes would be hugely helpful for identifying slides without exporting.

gog slides update-notes <presentationId> <slideId> --notes-file <path>

Update speaker notes on an existing slide without deleting/re-adding. Currently the only way to update notes is delete + re-add with the same image, which risks position errors.

gog slides replace-slide <presentationId> <slideId> <image> [--notes-file]

Atomic replace: swap the image (and optionally notes) on an existing slide in-place. This would be much safer than the current delete + add-before dance, which can go wrong if you pick the wrong anchor ID.

Context

These came up during a real editing session where 6 slides needed updating in an existing 39-slide deck. The delete/add-before workflow worked but caused two bugs:

• Deleted wrong slide because IDs are opaque
• Created a duplicate because I couldn't verify content before editing

The --before anchor approach also requires careful reverse-order processing when replacing multiple slides, since deletions shift subsequent positions.