Add mid-turn steering with Ctrl+S

cmd/root.go

@@ -811,7 +811,7 @@ func runNormalMode(ctx context.Context) error {
 			PrintError:    func(text string) { appInstance.PrintFromExtension("error", text) },
 			PrintBlock:    appInstance.PrintBlockFromExtension,
 			SendMessage:   func(text string) { appInstance.Run(text) },
-			CancelAndSend: func(text string) { appInstance.Steer(text) },
+			CancelAndSend: func(text string) { appInstance.InterruptAndSend(text) },
 			Exit:          func() { appInstance.QuitFromExtension() },
 			SetWidget: func(config extensions.WidgetConfig) {
 				kitInstance.SetExtensionWidget(config)

internal/agent/agent.go

@@ -275,7 +275,7 @@ func (a *Agent) GenerateWithLoopAndStreaming(ctx context.Context, messages []fan
 		var completedStepMessages []fantasy.Message
 
 		// Use fantasy's streaming agent
-		result, err := a.fantasyAgent.Stream(ctx, fantasy.AgentStreamCall{
+		streamCall := fantasy.AgentStreamCall{
 			Prompt:   prompt,
 			Files:    files,
 			Messages: history,
@@ -364,7 +364,51 @@ func (a *Agent) GenerateWithLoopAndStreaming(ctx context.Context, messages []fan
 				}
 				return nil
 			},
-		})
+		}
+
+		// If a steer channel is attached to the context, wire up a
+		// PrepareStep function that drains the channel between steps
+		// and injects pending steer messages as user messages before
+		// the next LLM call. This enables graceful mid-turn steering
+		// without cancelling in-progress tool execution.
+		if steerCh := steerChFromContext(ctx); steerCh != nil {
+			onConsumed := steerConsumedFromContext(ctx)
+			streamCall.PrepareStep = func(
+				stepCtx context.Context,
+				opts fantasy.PrepareStepFunctionOptions,
+			) (context.Context, fantasy.PrepareStepResult, error) {
+				// Drain all pending steer messages (non-blocking).
+				var steered []string
+				for {
+					select {
+					case msg := <-steerCh:
+						steered = append(steered, msg)
+					default:
+						goto done
+					}
+				}
+			done:
+				result := fantasy.PrepareStepResult{
+					Model:    opts.Model,
+					Messages: opts.Messages,
+				}
+				if len(steered) > 0 {
+					// Inject each steer message as a user message so the
+					// LLM sees the redirection on the next step.
+					for _, text := range steered {
+						result.Messages = append(result.Messages,
+							fantasy.NewUserMessage(text))
+					}
+					// Notify that steer messages were consumed.
+					if onConsumed != nil {
+						onConsumed(len(steered))
+					}
+				}
+				return stepCtx, result, nil
+			}
+		}
+
+		result, err := a.fantasyAgent.Stream(ctx, streamCall)
 		if err != nil {
 			// On cancellation (or any error), return a partial result
 			// containing messages from completed steps so the caller can

internal/agent/steer.go

@@ -0,0 +1,35 @@
+package agent
+
+import "context"
+
+// steerChKey is the context key for the steer channel.
+type steerChKey struct{}
+
+// steerConsumedKey is the context key for the steer-consumed callback.
+type steerConsumedKey struct{}
+
+// ContextWithSteerCh returns a new context with the steer channel attached.
+// The agent's PrepareStep function checks this channel between steps and
+// injects any pending steer messages as user messages before the next LLM call.
+func ContextWithSteerCh(ctx context.Context, ch <-chan string) context.Context {
+	return context.WithValue(ctx, steerChKey{}, ch)
+}
+
+// ContextWithSteerConsumed returns a new context with a callback that fires
+// when steer messages are consumed by PrepareStep. The count argument is the
+// number of messages injected in this batch.
+func ContextWithSteerConsumed(ctx context.Context, fn func(count int)) context.Context {
+	return context.WithValue(ctx, steerConsumedKey{}, fn)
+}
+
+// steerChFromContext extracts the steer channel from the context, or nil.
+func steerChFromContext(ctx context.Context) <-chan string {
+	ch, _ := ctx.Value(steerChKey{}).(<-chan string)
+	return ch
+}
+
+// steerConsumedFromContext extracts the steer-consumed callback, or nil.
+func steerConsumedFromContext(ctx context.Context) func(int) {
+	fn, _ := ctx.Value(steerConsumedKey{}).(func(int))
+	return fn
+}

internal/app/app.go

@@ -159,11 +159,57 @@ func (a *App) QueueLength() int {
 	return len(a.queue)
 }
 
-// Steer cancels the current agent step (if running), clears the queue, and
-// sends a new message that will execute as soon as the current step finishes
-// cancelling. If the agent is idle, the message executes immediately.
-// This is the "steer" delivery mode for SendMessage.
-func (a *App) Steer(prompt string) {
+// Steer injects a steering message into the currently running agent turn.
+// If the agent is in a multi-step tool loop, the message is delivered after
+// the current tool execution finishes but before the next LLM call (graceful
+// mid-turn injection via Fantasy's PrepareStep). If the agent is streaming
+// a text-only response (no pending tool calls), the message waits until the
+// response completes and then executes as the next turn.
+//
+// If the agent is idle, the message starts executing immediately (same as Run).
+//
+// Returns the number of pending steer/queue items (0 = started immediately,
+// >0 = injected/queued). The caller must update UI state based on the return
+// value — Steer does NOT send events to the program to avoid deadlocking
+// when called from within Update().
+//
+// Satisfies ui.AppController.
+func (a *App) Steer(prompt string) int {
+	a.mu.Lock()
+
+	if a.closed {
+		a.mu.Unlock()
+		return 0
+	}
+
+	if !a.busy {
+		// Not busy — start immediately, same as Run().
+		item := queueItem{Prompt: prompt}
+		a.busy = true
+		a.wg.Add(1)
+		a.mu.Unlock()
+		go a.drainQueue(item)
+		return 0
+	}
+
+	a.mu.Unlock()
+
+	// Agent is busy — inject via the SDK's steer channel. The message
+	// will be picked up by PrepareStep between agent steps (after tool
+	// execution, before next LLM call). If PrepareStep doesn't fire
+	// (text-only response), drainQueue will pick it up after the turn.
+	if a.opts.Kit != nil {
+		a.opts.Kit.InjectSteer(prompt)
+	}
+	return 1
+}
+
+// InterruptAndSend cancels the current agent step (if running), clears the
+// queue, and sends a new message that will execute as soon as the current
+// step finishes cancelling. If the agent is idle, the message executes
+// immediately. This is the hard-cancel delivery mode used by extensions'
+// CancelAndSend.
+func (a *App) InterruptAndSend(prompt string) {
 	a.mu.Lock()
 
 	if a.closed {
@@ -434,6 +480,24 @@ func (a *App) drainQueue(first queueItem) {
 		// Process all collected items as a single batch
 		a.runQueueBatch(items)
 
+		// Drain any unconsumed steer messages from the SDK channel.
+		// These arrive when the user steered during a text-only response
+		// (no tool calls, so PrepareStep didn't fire for a second step).
+		// They go to the front of the queue so they run next.
+		if a.opts.Kit != nil {
+			if leftover := a.opts.Kit.DrainSteer(); len(leftover) > 0 {
+				a.mu.Lock()
+				steerItems := make([]queueItem, len(leftover))
+				for i, text := range leftover {
+					steerItems[i] = queueItem{Prompt: text}
+				}
+				a.queue = append(steerItems, a.queue...)
+				a.mu.Unlock()
+				// Notify UI about the consumed steer messages.
+				a.sendEvent(SteerConsumedEvent{})
+			}
+		}
+
 		// Check if more items were queued while we were processing
 		a.mu.Lock()
 		hasMore := len(a.queue) > 0
@@ -687,6 +751,8 @@ func (a *App) subscribeSDKEvents(sendFn func(tea.Msg)) func() {
 					int(ev.CacheWriteTokens),
 				)
 			}
+		case kit.SteerConsumedEvent:
+			sendFn(SteerConsumedEvent{})
 		}
 	}))
 

internal/app/events.go

@@ -141,6 +141,12 @@ type CompactErrorEvent struct {
 	Err error
 }
 
+// SteerConsumedEvent is sent when one or more steering messages have been
+// consumed — either injected mid-turn via PrepareStep, or drained into the
+// queue after a turn completes. The TUI uses this to clear the steering
+// badge from the display.
+type SteerConsumedEvent struct{}
+
 // ModelChangedEvent is sent when an extension changes the active model via
 // ctx.SetModel. The TUI updates the model name shown in the status bar and
 // message attribution.

internal/ui/input.go

@@ -65,6 +65,10 @@ type InputComponent struct {
 	// hideHint suppresses the "enter submit · ctrl+j..." hint text.
 	hideHint bool
 
+	// agentBusy indicates the agent is currently working. When true, the
+	// hint text shows steering shortcut (Ctrl+S) instead of submit.
+	agentBusy bool
+
 	// pendingImages holds clipboard images attached to the next submission.
 	// Images are added via Ctrl+V and cleared on submit or Ctrl+U.
 	pendingImages []ImageAttachment
@@ -514,7 +518,16 @@ func (s *InputComponent) View() tea.View {
 		// Adapt hint text to available width (accounting for left padding of 3).
 		var hint string
 		availableHintWidth := s.width - 3
-		if availableHintWidth >= 67 {
+		if s.agentBusy {
+			// When the agent is working, show steering shortcut.
+			if availableHintWidth >= 55 {
+				hint = "enter queue • ctrl+s steer • esc esc cancel"
+			} else if availableHintWidth >= 35 {
+				hint = "↵ queue • ^S steer • esc×2 cancel"
+			} else {
+				hint = "^S steer"
+			}
+		} else if availableHintWidth >= 67 {
 			hint = "enter submit • ctrl+j / shift+enter new line • ctrl+v paste image"
 		} else if availableHintWidth >= 40 {
 			hint = "↵ submit • ctrl+j newline • ctrl+v image"

internal/ui/model.go

@@ -98,6 +98,12 @@ type AppController interface {
 	// alongside the text. Returns the current queue depth (0 = started
 	// immediately, >0 = queued).
 	RunWithFiles(prompt string, files []fantasy.FilePart) int
+	// Steer injects a steering message into the currently running agent
+	// turn. If the agent is busy, the message is delivered between steps
+	// (after current tool finishes, before next LLM call). If idle, the
+	// message starts executing immediately. Returns 0 if started
+	// immediately, >0 if injected/pending.
+	Steer(prompt string) int
 }
 
 // SkillItem holds display metadata about a loaded skill for the startup
@@ -415,6 +421,11 @@ type AppModel struct {
 	// the input and move to scrollback when the agent picks them up.
 	queuedMessages []string
 
+	// steeringMessages stores the text of prompts that were sent as steer
+	// messages (injected mid-turn via Ctrl+S). Rendered with a "STEERING"
+	// badge above the input. Cleared when the steer is consumed.
+	steeringMessages []string
+
 	// pendingUserPrints holds user messages that have been consumed from the
 	// queue but not yet printed to scrollback. They are deferred until
 	// SpinnerEvent{Show: true} so the previous assistant response can be
@@ -1070,6 +1081,45 @@ func (m *AppModel) Update(msg tea.Msg) (tea.Model, tea.Cmd) {
 				return m, tea.Batch(cmds...)
 			}
 			// In other states pass ESC through to children below.
+
+		case "ctrl+s":
+			// Steer: inject the current input as a steering message into the
+			// running agent turn. Only active during stateWorking — in input
+			// state, Ctrl+S is passed through to children (no-op by default).
+			if m.state == stateWorking && m.appCtrl != nil {
+				var text string
+				if ic, ok := m.input.(*InputComponent); ok {
+					text = strings.TrimSpace(ic.textarea.Value())
+				}
+				if text != "" {
+					// Clear the input and push to history.
+					if ic, ok := m.input.(*InputComponent); ok {
+						ic.pushHistory(text)
+						ic.textarea.SetValue("")
+					}
+
+					// Preprocess @file references.
+					processedText := text
+					if m.cwd != "" {
+						processedText = ProcessFileAttachments(text, m.cwd)
+					}
+
+					// Inject the steer message.
+					sLen := m.appCtrl.Steer(processedText)
+					if sLen > 0 {
+						m.steeringMessages = append(m.steeringMessages, text)
+						m.distributeHeight()
+					} else {
+						// Started immediately (agent was idle).
+						m.pendingUserPrints = append(m.pendingUserPrints, text)
+						m.flushStreamAndPendingUserMessages()
+						if m.state != stateWorking {
+							m.state = stateWorking
+						}
+					}
+				}
+				return m, tea.Batch(cmds...)
+			}
 		}
 
 		// Route key events to the focused child. Check for editor
@@ -1389,6 +1439,14 @@ func (m *AppModel) Update(msg tea.Msg) (tea.Model, tea.Cmd) {
 		}
 		m.distributeHeight()
 
+	case app.SteerConsumedEvent:
+		// Steering messages were consumed — either injected mid-turn
+		// via PrepareStep or drained into the queue after a turn.
+		// Move them to pendingUserPrints for scrollback rendering.
+		m.pendingUserPrints = append(m.pendingUserPrints, m.steeringMessages...)
+		m.steeringMessages = m.steeringMessages[:0]
+		m.distributeHeight()
+
 	case app.StepCompleteEvent:
 		// Keep stream content visible in the view — don't flush to scrollback
 		// yet. Flushing + resetting in the same frame would shrink the view
@@ -1641,6 +1699,7 @@ func (m *AppModel) View() tea.View {
 	// Propagate hint visibility to the input component before rendering.
 	if ic, ok := m.input.(*InputComponent); ok {
 		ic.hideHint = vis.HideInputHint
+		ic.agentBusy = m.state == stateWorking
 	}
 
 	// When a prompt is active, it replaces the input area for consistency
@@ -1935,16 +1994,26 @@ func (m *AppModel) cycleThinkingLevel() {
 	go func() { _ = SaveThinkingLevelPreference(next) }()
 }
 
-// renderSeparator renders the separator line with an optional queue count badge.
+// renderSeparator renders the separator line with an optional queue/steer count badge.
 func (m *AppModel) renderSeparator() string {
 	theme := GetTheme()
 	lineStyle := lipgloss.NewStyle().Foreground(theme.Muted)
 	queueLen := len(m.queuedMessages)
+	steerLen := len(m.steeringMessages)
 
-	if queueLen > 0 {
-		badge := lipgloss.NewStyle().
-			Foreground(theme.Secondary).
-			Render(fmt.Sprintf("%d queued", queueLen))
+	if steerLen > 0 || queueLen > 0 {
+		var parts []string
+		if steerLen > 0 {
+			parts = append(parts, lipgloss.NewStyle().
+				Foreground(theme.Warning).
+				Render(fmt.Sprintf("%d steering", steerLen)))
+		}
+		if queueLen > 0 {
+			parts = append(parts, lipgloss.NewStyle().
+				Foreground(theme.Secondary).
+				Render(fmt.Sprintf("%d queued", queueLen)))
+		}
+		badge := strings.Join(parts, " ")
 
 		// Fill the separator with dashes up to the badge.
 		dashWidth := max(m.width-lipgloss.Width(badge)-1, 0)
@@ -2043,27 +2112,47 @@ func (m *AppModel) renderHeaderFooter(getter func() *WidgetData) string {
 	return renderContentBlock(data.Text, m.width, opts...)
 }
 
-// renderQueuedMessages renders queued prompts as styled content blocks with a
-// "QUEUED" badge, anchored between the separator and input. Each message is
-// displayed in a bordered block matching the overall message styling.
+// renderQueuedMessages renders queued and steering prompts as styled content
+// blocks with badges, anchored between the separator and input. Steering
+// messages use a distinct "STEERING" badge to differentiate from queued ones.
 func (m *AppModel) renderQueuedMessages() string {
-	if len(m.queuedMessages) == 0 {
+	if len(m.queuedMessages) == 0 && len(m.steeringMessages) == 0 {
 		return ""
 	}
 	theme := GetTheme()
-	badge := CreateBadge("QUEUED", theme.Accent)
 
 	var blocks []string
-	for _, msg := range m.queuedMessages {
-		content := msg + "\n" + badge
-		rendered := renderContentBlock(
-			content,
-			m.width,
-			WithAlign(lipgloss.Left),
-			WithBorderColor(theme.Muted),
-		)
-		blocks = append(blocks, rendered)
+
+	// Render steering messages first (higher priority).
+	if len(m.steeringMessages) > 0 {
+		badge := CreateBadge("STEERING", theme.Warning)
+		for _, msg := range m.steeringMessages {
+			content := msg + "\n" + badge
+			rendered := renderContentBlock(
+				content,
+				m.width,
+				WithAlign(lipgloss.Left),
+				WithBorderColor(theme.Warning),
+			)
+			blocks = append(blocks, rendered)
+		}
 	}
+
+	// Render queued messages.
+	if len(m.queuedMessages) > 0 {
+		badge := CreateBadge("QUEUED", theme.Accent)
+		for _, msg := range m.queuedMessages {
+			content := msg + "\n" + badge
+			rendered := renderContentBlock(
+				content,
+				m.width,
+				WithAlign(lipgloss.Left),
+				WithBorderColor(theme.Muted),
+			)
+			blocks = append(blocks, rendered)
+		}
+	}
+
 	return strings.Join(blocks, "\n")
 }
 
@@ -2134,6 +2223,7 @@ func (m *AppModel) handleSlashCommand(sc *SlashCommand) tea.Cmd {
 			m.appCtrl.ClearQueue()
 		}
 		m.queuedMessages = m.queuedMessages[:0]
+		m.steeringMessages = m.steeringMessages[:0]
 		m.distributeHeight()
 
 	case "/tree":
@@ -2321,7 +2411,9 @@ func (m *AppModel) printHelpMessage() {
 		"- `!!command`: Run shell command, output excluded from LLM context\n\n" +
 		"**Keys:**\n" +
 		"- `Ctrl+C`: Exit at any time\n" +
-		"- `ESC` (x2): Cancel ongoing LLM generation\n\n" +
+		"- `ESC` (x2): Cancel ongoing LLM generation\n" +
+		"- `Ctrl+S`: Steer — redirect the agent mid-turn (injected between tool calls)\n" +
+		"- `Enter` (while working): Queue message for after the agent finishes\n\n" +
 		"You can also just type your message to chat with the AI assistant."
 	m.printSystemMessage(help)
 }

internal/ui/model_test.go

@@ -67,6 +67,11 @@ func (s *stubAppController) RunWithFiles(prompt string, _ []fantasy.FilePart) in
 	return s.queueLen
 }
 
+func (s *stubAppController) Steer(prompt string) int {
+	s.runCalls = append(s.runCalls, prompt)
+	return s.queueLen
+}
+
 // --------------------------------------------------------------------------
 // Stub child components
 // --------------------------------------------------------------------------

pkg/kit/events.go

@@ -42,6 +42,9 @@ const (
 	// EventToolOutput fires when a tool produces streaming output chunks.
 	EventToolOutput EventType = "tool_output"
 	EventStepUsage  EventType = "step_usage"
+	// EventSteerConsumed fires when one or more steering messages have been
+	// injected into the agent turn via PrepareStep.
+	EventSteerConsumed EventType = "steer_consumed"
 )
 
 // ---------------------------------------------------------------------------
@@ -276,6 +279,16 @@ type CompactionEvent struct {
 // EventType implements Event.
 func (e CompactionEvent) EventType() EventType { return EventCompaction }
 
+// SteerConsumedEvent fires when one or more steering messages have been
+// injected into the agent turn via PrepareStep. The Count indicates how
+// many messages were consumed in this batch.
+type SteerConsumedEvent struct {
+	Count int
+}
+
+// EventType implements Event.
+func (e SteerConsumedEvent) EventType() EventType { return EventSteerConsumed }
+
 // ---------------------------------------------------------------------------
 // EventBus
 // ---------------------------------------------------------------------------

pkg/kit/kit.go

@@ -66,6 +66,13 @@ type Kit struct {
 	// subagentListeners holds per-tool-call event listeners registered via
 	// SubscribeSubagent(). Keyed by toolCallID → *subagentListenerSet.
 	subagentListeners sync.Map
+
+	// steerCh is a buffered channel used to inject steering messages into
+	// the running agent turn via Fantasy's PrepareStep. Created fresh for
+	// each generate() call and set to nil when idle. Protected by steerMu.
+	steerMu       sync.Mutex
+	steerCh       chan string
+	leftoverSteer []string // unconsumed steer messages from the last turn
 }
 
 // Subscribe registers an EventListener that will be called for every lifecycle
@@ -1405,6 +1412,35 @@ func (m *Kit) Subagent(ctx context.Context, cfg SubagentConfig) (*SubagentResult
 // All prompt modes (Prompt, Steer, FollowUp, PromptWithOptions) share this
 // single code path so callback wiring is never duplicated.
 func (m *Kit) generate(ctx context.Context, messages []fantasy.Message) (*agent.GenerateWithLoopResult, error) {
+	// Create a per-turn steer channel and attach it to the context so the
+	// agent's PrepareStep can inject steering messages between steps.
+	steerCh := make(chan string, 16)
+	m.steerMu.Lock()
+	m.steerCh = steerCh
+	m.steerMu.Unlock()
+	defer func() {
+		// Drain any unconsumed steer messages before nilling the channel.
+		// These are stored in leftoverSteer so DrainSteer() can return them.
+		var leftover []string
+		for {
+			select {
+			case msg := <-steerCh:
+				leftover = append(leftover, msg)
+			default:
+				goto drained
+			}
+		}
+	drained:
+		m.steerMu.Lock()
+		m.steerCh = nil
+		m.leftoverSteer = leftover
+		m.steerMu.Unlock()
+	}()
+	ctx = agent.ContextWithSteerCh(ctx, steerCh)
+	ctx = agent.ContextWithSteerConsumed(ctx, func(count int) {
+		m.events.emit(SteerConsumedEvent{Count: count})
+	})
+
 	// Inject the in-process subagent spawner into the context so the
 	// spawn_subagent core tool can create child Kit instances without
 	// importing pkg/kit (which would create an import cycle).
@@ -1714,6 +1750,71 @@ func (m *Kit) FollowUp(ctx context.Context, text string) (string, error) {
 	return result.Response, nil
 }
 
+// InjectSteer sends a steering message into the currently active agent turn.
+// The message will be injected as a user message between steps (after the
+// current tool execution finishes, before the next LLM call). If no turn is
+// active the message is silently dropped — callers should check IsGenerating()
+// or use Prompt()/Steer() for idle-state messaging.
+//
+// InjectSteer is safe to call from any goroutine. Multiple calls queue
+// messages in order; all pending steer messages are drained and injected
+// together at the next step boundary.
+//
+// This is the preferred way to redirect an agent mid-turn without cancelling
+// in-progress tool execution.
+func (m *Kit) InjectSteer(message string) {
+	m.steerMu.Lock()
+	ch := m.steerCh
+	m.steerMu.Unlock()
+	if ch == nil {
+		return
+	}
+	select {
+	case ch <- message:
+	default:
+		// Channel full — extremely unlikely with buffer of 16, but don't block.
+	}
+}
+
+// IsGenerating returns true if an agent turn is currently in progress.
+// Use this to decide between InjectSteer (mid-turn) and Prompt (new turn).
+func (m *Kit) IsGenerating() bool {
+	m.steerMu.Lock()
+	defer m.steerMu.Unlock()
+	return m.steerCh != nil
+}
+
+// DrainSteer removes and returns all unconsumed steer messages. Called after
+// a turn completes so the app layer can process any steer messages that
+// arrived after the last PrepareStep fired (e.g. during a text-only response
+// with no tool calls, or after the agent finished its last step).
+func (m *Kit) DrainSteer() []string {
+	m.steerMu.Lock()
+	defer m.steerMu.Unlock()
+
+	// First check leftover messages saved when generate() returned.
+	if len(m.leftoverSteer) > 0 {
+		msgs := m.leftoverSteer
+		m.leftoverSteer = nil
+		return msgs
+	}
+
+	// If a turn is still active, drain from the live channel.
+	if m.steerCh != nil {
+		var msgs []string
+		for {
+			select {
+			case msg := <-m.steerCh:
+				msgs = append(msgs, msg)
+			default:
+				return msgs
+			}
+		}
+	}
+
+	return nil
+}
+
 // PromptOptions configures a single PromptWithOptions call.
 type PromptOptions struct {
 	// SystemMessage is prepended as a system message before the user prompt.