| name | gemini-browser-image |
| description | Drive the logged-in Gemini web app in a real browser to generate or edit images without API keys. Use when browser persistence, existing login state, uploads, or real Gemini page behavior matter more than direct API access. |
Gemini Browser Image
Use this skill for Gemini website image workflows, not API-based image generation.
Use this skill for
- text-to-image in Gemini Web
- edit or reference-image generation
- browser-login-dependent image work
- saving real local outputs from the Gemini UI
Mandatory workflow
- Reuse the logged-in browser profile.
- Normalize Gemini into a stable image-generation state before acting.
- Collect only minimum prompt/edit constraints.
- Submit, wait, and verify a real local file exists before declaring success.
Working rules
- One Gemini tab per job.
- Stop on CAPTCHA, login friction, or other human-verification walls.
- Do not claim success from a visible button alone; confirm a saved local artifact.
Read references as needed
references/upload-and-attach.md
For uploads and attachment behavior.references/output-capture.md
For save, verification, export, and fallback capture rules.
