Prompt History
Careti's prompt history works just like a terminal history (bash, zsh), letting you browse and reuse earlier messages. You no longer need to retype frequent commands or prompts, which greatly improves workflow speed.
How It Works
When the input field is focused, use the Up/Down arrow keys to navigate through previously entered messages.
Basic Usage
Up arrow: Move to the previous message. Keep pressing to go further back in time.
Down arrow: Move to the next message. If you pass the most recent message and press once more, it returns to your current in-progress text.
Example
1. Enter "Show me all TypeScript files in the project" and send
2. Enter "Read package.json" and send
3. Enter "Add usage instructions to README.md" and send
In the input field:
Up arrow once -> "Add usage instructions to README.md"
Up arrow again -> "Read package.json"
Up arrow again -> "Show me all TypeScript files in the project"
Down arrow once -> back to "Read package.json"
Key Features
Persistent Storage
Your inputs are saved automatically and persist across restarts. Even if you close and reopen a project, you can immediately recall past prompts.
History is stored per workspace, so prompts used in Project A are kept separate from Project B.
Automatic Deduplication
If you enter the exact same message consecutively, it is stored only once. For example, entering "run tests" three times in a row still records it just once, preventing noisy duplicates.
Not saved as duplicates:
"npm run test" entered
"npm run test" entered again -> not added to history
Saved as separate entries:
"npm run test" entered
"Show file list" entered
"npm run test" entered again -> added (because another command was in between)
Only Your Inputs Are Saved
History stores only what you typed. AI responses are not included, so arrow navigation stays clean and focused on your own prompts.
Up to 1000 Items
Careti stores up to 1000 history entries. When you exceed the limit, the oldest items are removed automatically to keep memory usage efficient.
For most workflows, 1000 entries is plenty for months of usage.
Consistent with the Terminal
Careti's prompt history behaves exactly like shell history in bash or zsh. The same Up/Down navigation and recall experience you already know in the terminal is available in Careti.
That consistency means you can use it immediately without learning a new interface.
Practical Tips
Automate Repetitive Work
Prompt history shines when you need to repeat similar tasks.
Scenario: apply the same pattern across multiple files
1. Enter "Add type annotations to all functions in UserController.ts"
2. After it finishes
3. Press Up arrow to recall the previous message
4. Replace "UserController" with "ProductController"
5. Send
Repeat across files quickly
Reuse Successful Prompts
If a prompt worked well before, you can quickly find and reuse it later.
Effective prompt used two weeks ago:
"Review this code. Analyze performance, readability, and security,
and list improvements with priorities."
When you want to reuse that pattern in another project:
Press Up arrow to browse back or locate the prompt from the same time period
Resume Projects Faster
When you reopen a project after a while, history helps you remember what you were working on.
Reopen a project after 2 weeks:
Press Up arrow to check recent history
-> "Add JWT refresh tokens to the auth system"
-> "Implement Redis session storage"
-> Oh, I was working on auth!
Keyboard Shortcut Reference
| Key | Action |
|---|---|
| Up Arrow | Move to previous message |
| Down Arrow | Move to next message |
| Enter | Send the currently displayed message |
| Esc | Cancel history navigation and return to current input |
Difference from Cline
Cline does not have prompt history. To reuse a previous message, you must retype it from scratch each time.
Careti's history system removes that friction and dramatically reduces repetitive input work.
FAQ
Q: Where is history stored?
A: It's stored automatically in VS Code's workspace storage. You don't need to manage files manually--VS Code handles it.
Q: Can I delete or reset history?
A: It is currently managed automatically. When the 1000-item limit is exceeded, older entries are removed, and if you delete the workspace, its history is deleted as well.
Q: Can I share history across projects?
A: No. History is managed per workspace, so each project keeps its own independent history. This helps preserve project-specific context.
Q: I entered sensitive information (API keys, etc.). Will it be stored?
A: Yes--everything you type is stored. Avoid placing sensitive data directly in prompts. Use environment variables or config files instead.
Q: Are multi-line inputs stored?
A: Yes. Multi-line messages are saved exactly as entered, including line breaks, and are restored when you recall them.
Prompt history is a Careti-exclusive feature. You can use the same Up/Down arrow navigation you're used to in terminals, directly in AI chat.