Ask Terminal AI

Ask Terminal AI is a command-line tool that allows users to quickly get and execute shell commands using AI or interact directly with AI in the terminal. It provides both a virtual terminal mode for command suggestions and a conversation mode for general queries.

⚡ DISCLAIMER: 99% of this project's code was generated by AI ⚡

Quickly download

1. Download file

wget -O ask "https://github.com/keizman/Ask-Terminal-AI/releases/download/main/ask_linux" 

2. Generate config.yaml and edit

If u have not config.yaml at conf dir, it will auto generate and hint you file location

ask 

Modify model and apikey

vim /root/.config/askta/config.yaml 

3. start to use

---

Quickly transfer

1. Copy from another machine

scp -r root@{ip}:/root/.config/askta /root/.config/

2. Download file

wget -O ask "https://github.com/keizman/Ask-Terminal-AI/releases/download/main/ask_linux" 

3. Make the binary executable:

chmod +x ask

4. Optionally install to the system path:

   sudo mv ask /usr/local/bin/ask

Building from Source

1. Clone the repository:

   git clone https://github.com/keizman/Ask-Terminal-AI.git
   # Or if you already have the code:
   cd Ask-Terminal-AI

2. Build the binary:

   go build -o ask main.go

3. Make the binary executable:

   chmod +x ask

4. Optionally install to the system path:

   sudo mv ask /usr/local/bin/ask

---

Configuration

Before using Ask Terminal AI, you need to create a configuration file.

1. Create a configuration directory:

   mkdir -p ~/.config/askta

2. Create a config.yaml file:

   nano ~/.config/askta/config.yaml

3. Add the following content (replace with your actual values):

   # AI service configuration
   base_url: "https://api.openai.com/v1/"  # API base URL
   api_key: "your-api-key"                 # API key
   model_name: "gpt-4o-mini"               # Model to use
   
   # Feature configuration
   private_mode: false                     # Privacy mode
   sys_prompt: "I'm using Linux"           # Custom system prompt
   
   # Provider configuration
   provider: "openai-compatible"           # Currently only supports openai-compatible

4. Save the file: Use Ctrl+O, press Enter, then Ctrl+X to exit nano.

---

Usage

Virtual Terminal Mode (Command Suggestions)

• To get AI-suggested commands, type:

  ask 

  Then, enter a query like:

  how to find the largest files on my system

• You'll get a list of suggested commands. Here are the key bindings:

  • Arrow keys (↑/↓): Navigate suggestions
  • Enter: Execute the selected command
  • Ctrl+q or Ctrl+C: Exit

---

Conversation Mode

• To have a conversation with the AI, use:

  ask -i "explain the difference between grep and awk"

• Or input file content

echo "Do you see spelling errors in the following text?"; cat script.sh | ask -i 

---

Options

Option	Description
-c, --config FILE	Specify configuration file location
-m, --model NAME	Temporarily specify model to use
-p, --provider NAME	Temporarily specify AI provider (currently only openai-compatible)
-u, --url URL	Temporarily specify API base URL
-k, --key KEY	Temporarily specify API key
-s, --sys-prompt TEXT	Temporarily specify system prompt
-x, --proxy URL	Specify proxy URL (e.g., http://user:pass@host:port)
--private-mode	Enable privacy mode
-v, --version	Show version information
-h, --help	Show help information
-show	Show command history

---

Example with Options

ask --model gpt-4 --sys-prompt "I'm using Ubuntu 22.04" "how to install Docker"

---

Logs

• Command history: C:\Users\{user}\AppData\Local\Temp\askta_Chistory.log
• Application logs: /tmp/askta_run.log

---

Security Notes

• API keys are stored encrypted on disk.
• Use --private-mode to avoid sending directory structure in queries.

---

Develop

generate go mod and sum

go mod init ask_terminal && go mod tidy

Build with stripping and optimization flags

go build -ldflags="-s -w" -o ask.exe  main.go

Use UPX to compress the binary (install UPX first)

upx --best ask.exe

(choco install upx)

Future Plans

• Support multi-turn conversations
• Support (i18n)
• ✓ Support proxy

License

Apache 2.0