AI Completions

AI chat, meeting analysis, and text completion commands

AI Completions

Commands

AI Chat Completions

60db ai:chat --prompt "Hello, how are you?"

With Custom System Prompt

60db ai:chat --prompt "Summarize this text" --system "You are a helpful assistant."

With Specific Model

60db ai:chat --prompt "Generate code" --model "qcall/slm-3b-int4" --temperature 0.7

AI Meeting Notes Analysis

60db ai:meeting --transcript "Meeting transcript here..." --title "Weekly Standup"

AI Text Completion

60db ai:complete --prompt "Once upon a time" --max-tokens 100

With JSON Output for Agents

60db --json ai:chat --prompt "Analyze this data"

Options

AI Chat Options

• -p, --prompt <text> - User prompt/message
• -m, --model <name> - Model name (default: qcall/slm-3b-int4)
• -s, --system <text> - System prompt
• --messages <json> - Messages as JSON array
• --max-tokens <number> - Max tokens (default: 2048)
• --temperature <number> - Temperature (default: 0.1)
• --top-k <number> - Top K (default: 5)
• --top-p <number> - Top P (default: 0.9)
• --stream <boolean> - Enable streaming (default: false)
• --enable-thinking <boolean> - Enable thinking (default: false)

AI Meeting Notes Options

• -p, --prompt <text> or -t, --transcript <text> - Meeting transcript
• --title <text> - Meeting title
• -m, --model <name> - Model name
• -s, --system <text> - Custom system prompt
• All other AI options supported

AI Text Completion Options

• -p, --prompt <text> - Prompt text (required)
• --max-tokens <number> - Max tokens to generate

Examples

AI Chat Completion

# Simple chat
60db ai:chat --prompt "Explain quantum computing"

# With system prompt for specific behavior
60db ai:chat --prompt "Review this code" --system "You are an expert code reviewer. Be concise and actionable."

# Higher temperature for creative responses
60db ai:chat --prompt "Write a poem about AI" --temperature 0.9

AI Meeting Notes Analysis

# Analyze meeting transcript
60db ai:meeting --transcript "John: We need to finish by Friday. Jane: I'll handle frontend." --title "Sprint Planning"

# Get structured JSON output
60db --json ai:meeting --transcript "$TRANSCRIPT" > meeting-notes.json