[Feature Request]: Save and Navigate Multiple Answer Variants within a Thread

[elter-tef]

Feature Description

Implement a feature within the Complexity extension to locally save and navigate between multiple answer variants generated for a single user prompt within a Perplexity thread. When a user edits their query or uses the "Rewrite" function, the extension should store the previous answer(s) in the browser's local storage (e.g., localStorage or IndexedDB). UI elements, such as navigation arrows (< >) or a variant counter (e.g., "2 of 3"), should be added alongside the answer display area, allowing the user to cycle through the locally saved variants for that specific conversational turn. It's understood that this data is stored only in the user's current browser session and Perplexity itself will only retain the latest response in the official thread history.

Use Case

This feature would significantly benefit users who frequently iterate on their prompts or use the "Rewrite" function to explore different perspectives or refine the AI's output. Currently, valuable previous responses are lost upon regeneration. This feature enables:

1. Direct Comparison: Users can easily compare different generated answers side-by-side (by cycling through them) without leaving the context of the current thread.
2. Tracking Refinement: Users can observe how changes in their prompt affect the AI's response, aiding in understanding prompt engineering.
3. Improved Workflow: Eliminates the need for cumbersome manual copying or creating separate threads for minor variations, keeping the conversation flow focused and organized within a single thread during an active session.
4. Enhanced Iteration: Makes the process of refining answers much smoother and more intuitive, similar to the user experience in platforms like ChatGPT or Open WebUI.

Ultimately, it enhances the immediate, in-browser interaction for users engaged in iterative query refinement.

Alternatives Considered

The current alternatives are manual and less efficient:

1. Manual Copy-Pasting: Users have to manually copy previous answers and paste them into external note-taking applications or documents before rewriting or editing the prompt. This is time-consuming and disrupts the workflow.
2. Creating New Threads: Users might create entirely new Perplexity threads for each variation of a prompt or after each rewrite. This fragments the conversation history, making it difficult to track the evolution of a single line of inquiry and clutters the user's thread list.