The pain
CLI agents are fast when the task is precise. UI feedback still collapses into long typed descriptions unless you mark the screen first.
How it works
- Hotkey Casso over the UI you care about.
- Draw numbered regions and optional notes.
- Send to write the local session and clipboard prompt.
- Paste into Grok CLI with the image reference and [n] notes.
Why Casso with Grok CLI
- Same desktop gesture as Claude Code and Codex - no Grok-only lock-in.
- Numbered multi-region feedback in one capture.
- Local-only sessions until you choose to paste.
- 7-day free trial, then $29 one-time personal license.
Workflow tips
If you bounce between Grok CLI and other agents in a week, an agent-agnostic overlay saves relearning. Casso's job is the map; the CLI agent's job is the edit.
For multi-step visual polish, re-annotate only what is still wrong after each pass. Fresh numbers beat a stale full-desktop shot with arrows from three turns ago.
FAQ
Related
Use Casso with Grok CLI
Numbered screen annotations for Claude Code, Codex, and other coding agents. Draw, prompt, paste - screenshots stay local. 7-day free trial, then $29 once.