1kiro-cli chat   # Start interactive session first, then use commands below

1/chat resume               # Interactive picker to resume a previous session
2/chat save /path/file.json # Save current session to a file
3/chat load /path/file.json # Load session from a file
4
5# Custom storage scripts
6/chat save-via-script ./scripts/save-to-git.sh   # Script receives JSON via stdin
7/chat load-via-script ./scripts/load-from-git.sh # Script outputs JSON to stdout

1/context show                  # Show context rules and matched files
2
3/context add src/app.js        # Add a specific file
4/context add "*.py"            # Add via glob pattern
5/context add "src/**/*.js"     # Recursive glob
6
7/context remove src/app.js     # Remove a rule
8/context clear                 # Remove all rules

1/model                            # Open interactive model picker
2/model claude-opus-4.6            # Select a model directly
3/model clau<Tab>                  # Tab-complete model names
4/model set-current-as-default     # Save current model as default

1/agent list                        # List all available agents
2/agent create my-agent             # Create a new agent
3/agent edit                        # Edit the current agent
4/agent edit my-agent               # Edit a specific agent
5/agent edit --path /abs/path.json  # Edit by file path
6/agent generate                    # AI-generate an agent config
7/agent schema                      # Show agent config JSON schema
8/agent set-default my-agent        # Set default agent for future sessions
9/agent swap code-reviewer          # Swap to a different agent mid-session

1/tools                  # View all tools, token counts, and permissions
2/tools schema           # Show input schema for all tools
3
4/tools trust write      # Trust a tool for this session (no confirmation)
5/tools untrust write    # Revert tool to per-request confirmation
6/tools trust-all        # Trust all tools
7/tools reset            # Reset all tools to default permission levels

1/prompts list                      # List all available prompts
2/prompts details code-review       # Show prompt details
3/prompts get code-review [arg]     # Retrieve a prompt
4@code-review [arg]                 # Shorthand for /prompts get
5
6/prompts create my-prompt          # Create a local prompt
7/prompts edit my-prompt            # Edit a local prompt
8/prompts remove my-prompt          # Delete a local prompt

 1/code init              # Initialize LSP-powered code intelligence
 2/code init -f           # Force reinitialize (restarts LSP servers)
 3/code overview          # Get workspace structure overview
 4/code overview --silent # Cleaner output
 5/code status            # Show LSP server + workspace status
 6
 7# Log viewing
 8/code logs                      # Last 20 ERROR logs
 9/code logs -l INFO              # INFO level and above
10/code logs -n 50                # Last 50 entries
11/code logs -l DEBUG -n 100      # Last 100 DEBUG+ logs
12/code logs -p ./lsp-logs.json   # Export logs to JSON

1/todos                           # View todo list (alias: /todo)
2/todo add "Fix authentication bug"
3/todo complete 1

1/mcp      # Show currently active MCP servers
2/hooks    # Display active context hooks for the session

1/compact     # Summarise conversation to free up context window space
2/paste       # Paste an image from clipboard into the conversation
3/editor      # Open $EDITOR to compose a long prompt
4/reply       # Open editor with last AI response quoted for reply
5/usage       # Show billing and credits information

1/experiment  # Toggle experimental features on/off
2/tangent     # Enter/exit tangent mode (isolated side conversation)
3             # Keyboard shortcut: Ctrl+T (when tangent mode is enabled)

1/issue    # Create a GitHub issue or feature request
2/logdump  # Create a zip of logs for support investigation