Prompt Runner for Google Gemini

This extension enables automatic prompt execution in the Google Gemini web app using additional URL parameters.

This extension allows automatic execution of prompts, which is not supported by default, by passing additional URL parameters to the Google Gemini web app. Features: - Automatic prompt execution via URL parameters (^1) - Replacement of the keyword `{{clipboard}}` in prompts with clipboard text (optional) - Automatic prompt submission (optional) - Prevent automatic execution when not logged into Gemini (optional) - Model selection via URL parameters - Generation of URLs usable in Prompt Runner for Google Gemini from the current prompt and selected model - Copying of the last response when `Ctrl + C` / `Cmd + C` is pressed without text selection (requires a copy button, so a certain window size is necessary) ^1: When the extension is installed, parameters are not sent to the server and are securely processed within the browser. Even if such a URL is shared with another environment, the parameters will not be sent to the server; however, if the extension is not installed in that environment, there is a risk that the parameters may be accessed by web analytics tools. Use Cases: - Register prompt templates in bookmarks - Trigger prompt execution from CLI Release Notes (ver1.6.2): - Fixed an issue where the extension could wait too long for Google Gemini to finish initializing. Release Notes (ver1.6.1): - Fixed an issue where the model switching button could not be found in the Google Gemini UI. Release Notes (ver1.6.0): - Added an options page to allow changing CSS selectors for identifying UI elements on the Gemini site, in preparation for site design changes. - Fixed an issue where the model switching button could not be found in the Google Gemini UI. Release Notes (ver1.5.3): - Fixed an issue where prompts were not executed automatically in some cases. - Fixed an issue where the copy button for responses could trigger the menu even when it was located outside the menu area. Release Notes (ver1.5.2): - Fixed an issue where the model switching button could not be found in the Google Gemini UI. Release Notes (ver1.5.1): - Fixed an issue where the model switching button could not be found in the Google Gemini UI. Release Notes (ver1.5.0): - Displayed a message recommending the fragment format for parameter specification over the query parameter format. - Updated the URL generation option to remember its selected state. - Improved the internal design. - Fixed an issue where the model could not be switched if the model name contained full-width parentheses unless it was an exact match. Release Notes (ver1.4.1): - Added support for fragment-style (#) parameter specification to improve privacy. - Introduced an experimental redirect page to prevent prompts from being sent to the server when the URL is opened in an environment without the extension installed. - Added an option to prevent automatic prompt execution when not logged into Gemini. - Fixed an issue where the UI could no longer be operated due to changes in Gemini’s interface. Release Notes (ver1.3.0): - Supports sending multiple prompts consecutively by combining the ext-send parameter with multiple ext-q parameters. - Updated the extension icon to display the progress of automated operations. Release Notes (ver1.2.0): - Added a feature to the popup that generates a usable URL for Prompt Runner for Google Gemini based on the current prompt and selected model. - Deprecated the `ext-confirm` parameter. Prompts will now only be auto-sent if explicitly allowed via the `ext-send` parameter. - Fixed an issue where the selected model was not being properly applied in some cases. Release Notes (ver1.1.0): - Added support for selecting models by name. - Improved the UI for opening the documentation, making it more intuitive. - Fixed a bug where cursor movement in the text area was affected when prompts included CR line breaks. Attribution: Google Gemini™ is a trademark of Google LLC.