fix(agent): track tool call args per ToolCallID for parallel calls (#33 )

Previously GenerateWithCallbacks stored the most recent tool call's args in a single shared variable, which got clobbered when a provider emitted multiple tool_use blocks in a single step. Every OnToolResult callback then received the args of the last OnToolCall, regardless of which call it was actually resolving — breaking any downstream UI, log, or trace that derived its description from the toolArgs parameter. - Replace the shared currentToolArgs with a map keyed by ToolCallID, guarded by a sync.Mutex in case the streaming layer dispatches callbacks from multiple goroutines. - Delete each entry in OnToolResult so the map cannot accumulate across steps. - Add a regression test driving the streaming wrapper with a fake fantasy.Agent that emits two parallel tool calls before either result, asserting each callback sees its own args. Fixes #33
chore(models): refresh embedded models.dev snapshot
2026-06-14 03:30:26 +00:00 · 2026-05-20 10:23:29 +03:00 · 2026-05-19 15:11:01 +03:00 · 2026-05-19 13:24:53 +03:00 · 2026-05-16 16:19:38 +03:00 · 2026-05-16 14:07:58 +03:00
32 changed files with 2195 additions and 550 deletions
@@ -1,268 +0,0 @@
-//go:build ignore
-
-package main
-
-import (
-	"context"
-	"encoding/json"
-	"fmt"
-	"os/exec"
-	"path/filepath"
-	"strings"
-	"time"
-
-	"kit/ext"
-)
-
-const (
-	diagnosticsTimeout = 20 * time.Second
-	maxOutputBytes     = 12_000
-)
-
-type toolPathInput struct {
-	Path string `json:"path"`
-}
-
-type lintResult struct {
-	Output string
-	Err    error
-}
-
-// Package-level state: set of .go files edited during the current agent turn.
-var editedFiles map[string]bool
-
-func Init(api ext.API) {
-	api.OnSessionStart(func(_ ext.SessionStartEvent, ctx ext.Context) {
-		ctx.Print("go-edit-lint extension loaded - will run gopls and golangci-lint after agent turns that edit Go files")
-	})
-
-	// Track edited .go files — don't lint yet.
-	api.OnToolResult(func(e ext.ToolResultEvent, ctx ext.Context) *ext.ToolResultResult {
-		if e.IsError || !isEditOrWrite(e.ToolName) {
-			return nil
-		}
-
-		absPath, ok := resolveGoFilePath(e.Input, ctx.CWD)
-		if !ok {
-			return nil
-		}
-
-		if editedFiles == nil {
-			editedFiles = make(map[string]bool)
-		}
-		editedFiles[absPath] = true
-		return nil
-	})
-
-	// After the agent turn ends, lint all collected files.
-	api.OnAgentEnd(func(e ext.AgentEndEvent, ctx ext.Context) {
-		if len(editedFiles) == 0 {
-			return
-		}
-
-		// Snapshot and reset immediately so the next turn starts clean.
-		files := editedFiles
-		editedFiles = nil
-
-		// Skip lint on errored turns.
-		if e.StopReason == "error" {
-			return
-		}
-
-		// Collect unique directories and file list for gopls.
-		var allGoplsOutput []string
-		for absPath := range files {
-			res := runGopls(ctx.CWD, absPath)
-			formatted := formatToolResult(res, "")
-			if formatted != "" {
-				allGoplsOutput = append(allGoplsOutput, fmt.Sprintf("# %s\n%s", filepath.Base(absPath), formatted))
-			}
-		}
-
-		lintRes := runGolangCILint(ctx.CWD, "./...")
-
-		goplsSection := "No diagnostics."
-		if len(allGoplsOutput) > 0 {
-			goplsSection = strings.Join(allGoplsOutput, "\n\n")
-		}
-		lintSection := formatToolResult(lintRes, "No lint issues.")
-
-		// Build file list for the report header.
-		var fileNames []string
-		for absPath := range files {
-			fileNames = append(fileNames, filepath.Base(absPath))
-		}
-
-		report := fmt.Sprintf(
-			"<go_diagnostics files=%q>\n[gopls]\n%s\n\n[golangci-lint]\n%s\n</go_diagnostics>",
-			strings.Join(fileNames, ", "),
-			goplsSection,
-			lintSection,
-		)
-
-		goplsIssues, lintIssues := countIssues(report)
-		hasIssues := goplsIssues > 0 || lintIssues > 0
-
-		if hasIssues {
-			// Show TUI block so the user sees it too.
-			var msgLines []string
-			msgLines = append(msgLines, fmt.Sprintf("Files: %s", strings.Join(fileNames, ", ")))
-			if goplsIssues > 0 {
-				msgLines = append(msgLines, fmt.Sprintf("gopls: %d issue(s)", goplsIssues))
-			}
-			if lintIssues > 0 {
-				msgLines = append(msgLines, fmt.Sprintf("golangci-lint: %d issue(s)", lintIssues))
-			}
-
-			borderColor := "#f9e2af" // yellow
-			if goplsIssues > 0 && lintIssues > 0 {
-				borderColor = "#f38ba8" // red
-			}
-
-			ctx.PrintBlock(ext.PrintBlockOpts{
-				Text:        strings.Join(msgLines, "\n"),
-				BorderColor: borderColor,
-				Subtitle:    "go-edit-lint",
-			})
-
-			// Inject a follow-up message so the agent fixes the issues.
-			ctx.SendMessage(report + "\n\n⚠️ DIAGNOSTICS FOUND: Please review and fix the issues above.")
-		} else {
-			ctx.PrintBlock(ext.PrintBlockOpts{
-				Text:        fmt.Sprintf("Files: %s\n✓ All clean", strings.Join(fileNames, ", ")),
-				BorderColor: "#a6e3a1",
-				Subtitle:    "go-edit-lint",
-			})
-		}
-	})
-}
-
-func isEditOrWrite(toolName string) bool {
-	return strings.EqualFold(toolName, "edit") || strings.EqualFold(toolName, "write")
-}
-
-func resolveGoFilePath(inputJSON, cwd string) (string, bool) {
-	var args toolPathInput
-	if err := json.Unmarshal([]byte(inputJSON), &args); err != nil || args.Path == "" {
-		return "", false
-	}
-
-	absPath := args.Path
-	if !filepath.IsAbs(absPath) {
-		absPath = filepath.Join(cwd, absPath)
-	}
-
-	if strings.ToLower(filepath.Ext(absPath)) != ".go" {
-		return "", false
-	}
-
-	return absPath, true
-}
-
-func runGopls(cwd, absPath string) lintResult {
-	ctx, cancel := context.WithTimeout(context.Background(), diagnosticsTimeout)
-	defer cancel()
-
-	cmd := exec.CommandContext(ctx, "gopls", "check", absPath)
-	cmd.Dir = cwd
-	out, err := cmd.CombinedOutput()
-
-	if ctx.Err() == context.DeadlineExceeded {
-		return lintResult{Err: fmt.Errorf("timed out after %s", diagnosticsTimeout)}
-	}
-
-	if err != nil {
-		return lintResult{Output: truncate(string(out), maxOutputBytes), Err: fmt.Errorf("failed to run gopls check: %w", err)}
-	}
-
-	return lintResult{Output: truncate(string(out), maxOutputBytes)}
-}
-
-func runGolangCILint(cwd, target string) lintResult {
-	ctx, cancel := context.WithTimeout(context.Background(), diagnosticsTimeout)
-	defer cancel()
-
-	args := []string{
-		"run",
-		target,
-		"--show-stats=false",
-		"--output.text.path", "stdout",
-		"--output.text.colors=false",
-		"--output.text.print-issued-lines=false",
-	}
-	cmd := exec.CommandContext(ctx, "golangci-lint", args...)
-	cmd.Dir = cwd
-	out, err := cmd.CombinedOutput()
-
-	if ctx.Err() == context.DeadlineExceeded {
-		return lintResult{Err: fmt.Errorf("timed out after %s", diagnosticsTimeout)}
-	}
-
-	trimmed := truncate(string(out), maxOutputBytes)
-	if err == nil {
-		return lintResult{Output: trimmed}
-	}
-
-	exitErr, ok := err.(*exec.ExitError)
-	if ok && exitErr.ExitCode() == 1 {
-		return lintResult{Output: trimmed}
-	}
-
-	return lintResult{Output: trimmed, Err: fmt.Errorf("failed to run golangci-lint: %w", err)}
-}
-
-func formatToolResult(res lintResult, emptyFallback string) string {
-	var lines []string
-	if res.Err != nil {
-		lines = append(lines, "ERROR: "+res.Err.Error())
-	}
-	out := strings.TrimSpace(res.Output)
-	if out == "" {
-		if res.Err == nil {
-			if emptyFallback != "" {
-				lines = append(lines, emptyFallback)
-			}
-		}
-	} else {
-		lines = append(lines, out)
-	}
-	if len(lines) == 0 {
-		return emptyFallback
-	}
-	return strings.Join(lines, "\n")
-}
-
-func truncate(s string, max int) string {
-	if len(s) <= max {
-		return s
-	}
-	return s[:max] + "\n... output truncated ..."
-}
-
-func countIssues(report string) (goplsCount, lintCount int) {
-	goplsStart := strings.Index(report, "[gopls]")
-	lintStart := strings.Index(report, "[golangci-lint]")
-	endTag := strings.Index(report, "</go_diagnostics>")
-
-	if goplsStart != -1 && lintStart != -1 {
-		goplsSection := report[goplsStart:lintStart]
-		for _, line := range strings.Split(goplsSection, "\n") {
-			line = strings.TrimSpace(line)
-			if line != "" && line != "[gopls]" && line != "No diagnostics." && !strings.HasPrefix(line, "#") {
-				goplsCount++
-			}
-		}
-	}
-
-	if lintStart != -1 && endTag != -1 {
-		lintSection := report[lintStart:endTag]
-		for _, line := range strings.Split(lintSection, "\n") {
-			line = strings.TrimSpace(line)
-			if line != "" && line != "[golangci-lint]" && line != "No lint issues." {
-				lintCount++
-			}
-		}
-	}
-
-	return goplsCount, lintCount
-}
@@ -434,6 +434,36 @@ func extensionCommandsForUI(k *kit.Kit) []commands.ExtensionCommand {
 	return cmds
 }

+// buildExtensionItems converts the loaded extensions into ui.ExtensionItem
+// values for the startup [Extensions] section. The display name is the file
+// basename without the .go suffix; for subdirectory extensions the parent
+// directory name is used (matching `kit extensions list`).
+func buildExtensionItems(k *kit.Kit, cwd string) []ui.ExtensionItem {
+	infos := k.Extensions().Loaded()
+	if len(infos) == 0 {
+		return nil
+	}
+	items := make([]ui.ExtensionItem, 0, len(infos))
+	for _, info := range infos {
+		name := filepath.Base(info.Path)
+		if name == "main.go" {
+			// Subdirectory extension: use the parent directory's name.
+			name = filepath.Base(filepath.Dir(info.Path))
+		}
+		name = strings.TrimSuffix(name, ".go")
+		source := "user"
+		if cwd != "" && strings.HasPrefix(info.Path, cwd) {
+			source = "project"
+		}
+		items = append(items, ui.ExtensionItem{
+			Name:   name,
+			Path:   info.Path,
+			Source: source,
+		})
+	}
+	return items
+}
+
 // widgetProviderForUI returns a function that converts extension widgets to
 // ui.WidgetData for the given placement. Returns nil if extensions are
 // disabled, which is safe — the UI treats a nil GetWidgets as "no widgets".
@@ -784,6 +814,16 @@ func runNormalMode(ctx context.Context) error {
 	}
 	defer func() { _ = kitInstance.Close() }()

+	// Build the "System Prompt loaded" notice shown at startup, paralleling the
+	// per-server "MCP server loaded" notifications so users can confirm that a
+	// configured prompt file was found and applied.
+	var systemPromptLoadedMsg string
+	if kitInstance.HasCustomSystemPrompt() {
+		if src := kitInstance.GetSystemPromptSource(); src != "" {
+			systemPromptLoadedMsg = "System Prompt loaded: " + src
+		}
+	}
+
 	// Extract metadata for display and app options.
 	parsedProvider, modelName, serverNames, toolNames, mcpToolCount, extensionToolCount := CollectAgentMetadata(kitInstance, mcpConfig)

@@ -801,6 +841,9 @@ func runNormalMode(ctx context.Context) error {
 		}

 		DisplayDebugConfig(cli, kitInstance, mcpConfig, parsedProvider)
+		if systemPromptLoadedMsg != "" && cli != nil {
+			cli.DisplayInfo(systemPromptLoadedMsg)
+		}
 	}

 	// Load existing messages from resumed/continued sessions.
@@ -840,6 +883,9 @@ func runNormalMode(ctx context.Context) error {

 	// Buffer for extension messages during startup (printed after startup banner).
 	var startupExtensionMessages []string
+	if systemPromptLoadedMsg != "" {
+		startupExtensionMessages = append(startupExtensionMessages, systemPromptLoadedMsg)
+	}

 	// Set up extension context and emit SessionStart.
 	if kitInstance.Extensions().HasExtensions() {
@@ -919,12 +965,17 @@ func runNormalMode(ctx context.Context) error {
 			source = "project"
 		}
 		skillItems = append(skillItems, ui.SkillItem{
-			Name:   s.Name,
-			Path:   s.Path,
-			Source: source,
+			Name:        s.Name,
+			Path:        s.Path,
+			Source:      source,
+			Description: s.Description,
 		})
 	}

+	// Build extension items from the loaded extensions for the [Extensions]
+	// startup section. Each entry is a single .go file (or a subdir's main.go).
+	extensionItems := buildExtensionItems(kitInstance, cwd)
+
 	// Build prompt template and skill item provider callbacks for hot-reload.
 	// These are called by the TUI when ContentReloadEvent fires.
 	getPromptTemplates := func() []*prompts.PromptTemplate {
@@ -960,14 +1011,22 @@ func runNormalMode(ctx context.Context) error {
 				source = "project"
 			}
 			items = append(items, ui.SkillItem{
-				Name:   s.Name,
-				Path:   s.Path,
-				Source: source,
+				Name:        s.Name,
+				Path:        s.Path,
+				Source:      source,
+				Description: s.Description,
 			})
 		}
 		return items
 	}

+	// getExtensionItems re-collects the loaded extension list, used by the
+	// TUI after an extension hot-reload to refresh the [Extensions] row.
+	getExtensionItems := func() []ui.ExtensionItem {
+		cwd, _ := os.Getwd()
+		return buildExtensionItems(kitInstance, cwd)
+	}
+
 	// Build extension UI providers once (shared between both modes).
 	getWidgets := widgetProviderForUI(kitInstance)
 	getHeader := headerProviderForUI(kitInstance)
@@ -1147,7 +1206,10 @@ func runNormalMode(ctx context.Context) error {
 		extWatcher, watchErr := extensions.NewWatcher(watchDirs, func() {
 			if err := reloadExtensionsForUI(); err != nil {
 				log.Printf("auto-reload extensions failed: %v", err)
+				appInstance.PrintFromExtension("error", fmt.Sprintf("Extension auto-reload failed: %v", err))
+				return
 			}
+			appInstance.PrintFromExtension("info", "Extensions reloaded.")
 		})
 		if watchErr != nil {
 			log.Printf("extension file watcher not started: %v", watchErr)
@@ -1166,6 +1228,7 @@ func runNormalMode(ctx context.Context) error {
 		promptDirs := watcher.CollectDirs(
 			[]string{
 				filepath.Join(homeDir, ".kit", "prompts"),
+				prompts.GlobalDir(),
 				filepath.Join(cwd, ".kit", "prompts"),
 			},
 			append(promptTemplatePaths, viper.GetStringSlice("prompts")...),
@@ -1204,7 +1267,7 @@ func runNormalMode(ctx context.Context) error {

 	// Check if running in non-interactive mode
 	if positionalPrompt != "" {
-		return runNonInteractiveModeApp(ctx, appInstance, cli, positionalPrompt, quietFlag, jsonFlag, noExitFlag, modelName, parsedProvider, kitInstance.GetLoadingMessage(), serverNames, toolNames, mcpToolCount, extensionToolCount, usageTracker, extCommands, promptTemplates, contextPaths, skillItems, getPromptTemplates, getSkillItems, getToolNames, getMCPToolCount, mcpPrompts, getMCPPrompts, expandMCPPrompt, getWidgets, getHeader, getFooter, getToolRenderer, getEditorInterceptor, getUIVisibility, getStatusBarEntries, emitBeforeFork, emitBeforeSessionSwitch, getGlobalShortcuts, getExtensionCommands, setModelForUI, emitModelChangeForUI, kitInstance.IsReasoningModel(), kitInstance.GetThinkingLevel(), setThinkingLevelForUI, switchSessionForUI, reloadExtensionsForUI)
+		return runNonInteractiveModeApp(ctx, appInstance, cli, positionalPrompt, quietFlag, jsonFlag, noExitFlag, modelName, parsedProvider, kitInstance.GetLoadingMessage(), serverNames, toolNames, mcpToolCount, extensionToolCount, usageTracker, extCommands, promptTemplates, contextPaths, skillItems, extensionItems, getPromptTemplates, getSkillItems, getExtensionItems, getToolNames, getMCPToolCount, mcpPrompts, getMCPPrompts, expandMCPPrompt, getWidgets, getHeader, getFooter, getToolRenderer, getEditorInterceptor, getUIVisibility, getStatusBarEntries, emitBeforeFork, emitBeforeSessionSwitch, getGlobalShortcuts, getExtensionCommands, setModelForUI, emitModelChangeForUI, kitInstance.IsReasoningModel(), kitInstance.GetThinkingLevel(), setThinkingLevelForUI, switchSessionForUI, reloadExtensionsForUI)
 	}

 	// Quiet mode is not allowed in interactive mode
@@ -1212,7 +1275,7 @@ func runNormalMode(ctx context.Context) error {
 		return fmt.Errorf("--quiet requires a prompt")
 	}

-	return runInteractiveModeBubbleTea(ctx, appInstance, modelName, parsedProvider, kitInstance.GetLoadingMessage(), serverNames, toolNames, mcpToolCount, extensionToolCount, usageTracker, extCommands, promptTemplates, contextPaths, skillItems, getPromptTemplates, getSkillItems, getToolNames, getMCPToolCount, mcpPrompts, getMCPPrompts, expandMCPPrompt, getWidgets, getHeader, getFooter, getToolRenderer, getEditorInterceptor, getUIVisibility, getStatusBarEntries, emitBeforeFork, emitBeforeSessionSwitch, getGlobalShortcuts, getExtensionCommands, setModelForUI, emitModelChangeForUI, kitInstance.IsReasoningModel(), kitInstance.GetThinkingLevel(), setThinkingLevelForUI, switchSessionForUI, reloadExtensionsForUI, startupExtensionMessages)
+	return runInteractiveModeBubbleTea(ctx, appInstance, modelName, parsedProvider, kitInstance.GetLoadingMessage(), serverNames, toolNames, mcpToolCount, extensionToolCount, usageTracker, extCommands, promptTemplates, contextPaths, skillItems, extensionItems, getPromptTemplates, getSkillItems, getExtensionItems, getToolNames, getMCPToolCount, mcpPrompts, getMCPPrompts, expandMCPPrompt, getWidgets, getHeader, getFooter, getToolRenderer, getEditorInterceptor, getUIVisibility, getStatusBarEntries, emitBeforeFork, emitBeforeSessionSwitch, getGlobalShortcuts, getExtensionCommands, setModelForUI, emitModelChangeForUI, kitInstance.IsReasoningModel(), kitInstance.GetThinkingLevel(), setThinkingLevelForUI, switchSessionForUI, reloadExtensionsForUI, startupExtensionMessages)
 }

 // runNonInteractiveModeApp executes a single prompt via the app layer and exits,
@@ -1225,7 +1288,7 @@ func runNormalMode(ctx context.Context) error {
 //
 // When --no-exit is set, after the prompt completes the interactive BubbleTea
 // TUI is started so the user can continue the conversation.
-func runNonInteractiveModeApp(ctx context.Context, appInstance *app.App, cli *ui.CLI, prompt string, quiet, jsonOutput, noExit bool, modelName, providerName, loadingMessage string, serverNames, toolNames []string, mcpToolCount, extensionToolCount int, usageTracker *ui.UsageTracker, extCommands []commands.ExtensionCommand, promptTemplates []*prompts.PromptTemplate, contextPaths []string, skillItems []ui.SkillItem, getPromptTemplates func() []*prompts.PromptTemplate, getSkillItems func() []ui.SkillItem, getToolNames func() []string, getMCPToolCount func() int, mcpPrompts []ui.MCPPromptInfo, getMCPPrompts func() []ui.MCPPromptInfo, expandMCPPrompt func(string, string, map[string]string) (*ui.MCPPromptExpandResult, error), getWidgets func(string) []ui.WidgetData, getHeader, getFooter func() *ui.WidgetData, getToolRenderer func(string) *ui.ToolRendererData, getEditorInterceptor func() *ui.EditorInterceptor, getUIVisibility func() *ui.UIVisibility, getStatusBarEntries func() []ui.StatusBarEntryData, emitBeforeFork func(string, bool, string) (bool, string), emitBeforeSessionSwitch func(string) (bool, string), getGlobalShortcuts func() map[string]func(), getExtensionCommands func() []commands.ExtensionCommand, setModel func(string) error, emitModelChange func(string, string, string), isReasoningModel bool, thinkingLevel string, setThinkingLevel func(string) error, switchSession func(string) error, reloadExtensions func() error) error {
+func runNonInteractiveModeApp(ctx context.Context, appInstance *app.App, cli *ui.CLI, prompt string, quiet, jsonOutput, noExit bool, modelName, providerName, loadingMessage string, serverNames, toolNames []string, mcpToolCount, extensionToolCount int, usageTracker *ui.UsageTracker, extCommands []commands.ExtensionCommand, promptTemplates []*prompts.PromptTemplate, contextPaths []string, skillItems []ui.SkillItem, extensionItems []ui.ExtensionItem, getPromptTemplates func() []*prompts.PromptTemplate, getSkillItems func() []ui.SkillItem, getExtensionItems func() []ui.ExtensionItem, getToolNames func() []string, getMCPToolCount func() int, mcpPrompts []ui.MCPPromptInfo, getMCPPrompts func() []ui.MCPPromptInfo, expandMCPPrompt func(string, string, map[string]string) (*ui.MCPPromptExpandResult, error), getWidgets func(string) []ui.WidgetData, getHeader, getFooter func() *ui.WidgetData, getToolRenderer func(string) *ui.ToolRendererData, getEditorInterceptor func() *ui.EditorInterceptor, getUIVisibility func() *ui.UIVisibility, getStatusBarEntries func() []ui.StatusBarEntryData, emitBeforeFork func(string, bool, string) (bool, string), emitBeforeSessionSwitch func(string) (bool, string), getGlobalShortcuts func() map[string]func(), getExtensionCommands func() []commands.ExtensionCommand, setModel func(string) error, emitModelChange func(string, string, string), isReasoningModel bool, thinkingLevel string, setThinkingLevel func(string) error, switchSession func(string) error, reloadExtensions func() error) error {
 	// Expand @file references in the prompt before sending to the agent.
 	// Text files are XML-inlined; binary files are extracted as multimodal parts.
 	var fileParts []kit.LLMFilePart
@@ -1286,7 +1349,7 @@ func runNonInteractiveModeApp(ctx context.Context, appInstance *app.App, cli *ui

 	// If --no-exit was requested, hand off to the interactive TUI.
 	if noExit {
-		return runInteractiveModeBubbleTea(ctx, appInstance, modelName, providerName, loadingMessage, serverNames, toolNames, mcpToolCount, extensionToolCount, usageTracker, extCommands, promptTemplates, contextPaths, skillItems, getPromptTemplates, getSkillItems, getToolNames, getMCPToolCount, mcpPrompts, getMCPPrompts, expandMCPPrompt, getWidgets, getHeader, getFooter, getToolRenderer, getEditorInterceptor, getUIVisibility, getStatusBarEntries, emitBeforeFork, emitBeforeSessionSwitch, getGlobalShortcuts, getExtensionCommands, setModel, emitModelChange, isReasoningModel, thinkingLevel, setThinkingLevel, switchSession, reloadExtensions, nil)
+		return runInteractiveModeBubbleTea(ctx, appInstance, modelName, providerName, loadingMessage, serverNames, toolNames, mcpToolCount, extensionToolCount, usageTracker, extCommands, promptTemplates, contextPaths, skillItems, extensionItems, getPromptTemplates, getSkillItems, getExtensionItems, getToolNames, getMCPToolCount, mcpPrompts, getMCPPrompts, expandMCPPrompt, getWidgets, getHeader, getFooter, getToolRenderer, getEditorInterceptor, getUIVisibility, getStatusBarEntries, emitBeforeFork, emitBeforeSessionSwitch, getGlobalShortcuts, getExtensionCommands, setModel, emitModelChange, isReasoningModel, thinkingLevel, setThinkingLevel, switchSession, reloadExtensions, nil)
 	}

 	return nil
@@ -1384,7 +1447,7 @@ func writeJSONError(err error) {
 //  4. Calls program.Run() which blocks until the user quits (Ctrl+C or /quit).
 //
 // SetupCLI is not used for interactive mode; the TUI (AppModel) handles its own rendering.
-func runInteractiveModeBubbleTea(_ context.Context, appInstance *app.App, modelName, providerName, loadingMessage string, serverNames, toolNames []string, mcpToolCount, extensionToolCount int, usageTracker *ui.UsageTracker, extCommands []commands.ExtensionCommand, promptTemplates []*prompts.PromptTemplate, contextPaths []string, skillItems []ui.SkillItem, getPromptTemplates func() []*prompts.PromptTemplate, getSkillItems func() []ui.SkillItem, getToolNames func() []string, getMCPToolCount func() int, mcpPrompts []ui.MCPPromptInfo, getMCPPrompts func() []ui.MCPPromptInfo, expandMCPPrompt func(string, string, map[string]string) (*ui.MCPPromptExpandResult, error), getWidgets func(string) []ui.WidgetData, getHeader, getFooter func() *ui.WidgetData, getToolRenderer func(string) *ui.ToolRendererData, getEditorInterceptor func() *ui.EditorInterceptor, getUIVisibility func() *ui.UIVisibility, getStatusBarEntries func() []ui.StatusBarEntryData, emitBeforeFork func(string, bool, string) (bool, string), emitBeforeSessionSwitch func(string) (bool, string), getGlobalShortcuts func() map[string]func(), getExtensionCommands func() []commands.ExtensionCommand, setModel func(string) error, emitModelChange func(string, string, string), isReasoningModel bool, thinkingLevel string, setThinkingLevel func(string) error, switchSession func(string) error, reloadExtensions func() error, startupExtensionMessages []string) error {
+func runInteractiveModeBubbleTea(_ context.Context, appInstance *app.App, modelName, providerName, loadingMessage string, serverNames, toolNames []string, mcpToolCount, extensionToolCount int, usageTracker *ui.UsageTracker, extCommands []commands.ExtensionCommand, promptTemplates []*prompts.PromptTemplate, contextPaths []string, skillItems []ui.SkillItem, extensionItems []ui.ExtensionItem, getPromptTemplates func() []*prompts.PromptTemplate, getSkillItems func() []ui.SkillItem, getExtensionItems func() []ui.ExtensionItem, getToolNames func() []string, getMCPToolCount func() int, mcpPrompts []ui.MCPPromptInfo, getMCPPrompts func() []ui.MCPPromptInfo, expandMCPPrompt func(string, string, map[string]string) (*ui.MCPPromptExpandResult, error), getWidgets func(string) []ui.WidgetData, getHeader, getFooter func() *ui.WidgetData, getToolRenderer func(string) *ui.ToolRendererData, getEditorInterceptor func() *ui.EditorInterceptor, getUIVisibility func() *ui.UIVisibility, getStatusBarEntries func() []ui.StatusBarEntryData, emitBeforeFork func(string, bool, string) (bool, string), emitBeforeSessionSwitch func(string) (bool, string), getGlobalShortcuts func() map[string]func(), getExtensionCommands func() []commands.ExtensionCommand, setModel func(string) error, emitModelChange func(string, string, string), isReasoningModel bool, thinkingLevel string, setThinkingLevel func(string) error, switchSession func(string) error, reloadExtensions func() error, startupExtensionMessages []string) error {
 	// Redirect all log output (stdlib and charm) to a file so that log
 	// messages don't write to stderr and corrupt the TUI. Bubble Tea
 	// captures stdout for rendering; any stray stderr output from
@@ -1429,6 +1492,8 @@ func runInteractiveModeBubbleTea(_ context.Context, appInstance *app.App, modelN
 		ContextPaths:             contextPaths,
 		SkillItems:               skillItems,
 		GetSkillItems:            getSkillItems,
+		ExtensionItems:           extensionItems,
+		GetExtensionItems:        getExtensionItems,
 		StartupExtensionMessages: startupExtensionMessages,
 		GetWidgets:               getWidgets,
 		GetHeader:                getHeader,
@@ -13,7 +13,7 @@ import (
 // without panicking and properly guards nil ctx calls.
 func TestSubagentMonitor_SessionStart(t *testing.T) {
 	harness := test.New(t)
-	harness.LoadFile("../../.kit/extensions/subagent-monitor.go")
+	harness.LoadFile("./subagent-monitor.go")

 	// Emit SessionStart - should not panic even with nil ctx functions
 	_, err := harness.Emit(extensions.SessionStartEvent{SessionID: "test-session"})
@@ -26,7 +26,7 @@ func TestSubagentMonitor_SessionStart(t *testing.T) {
 // creates entries and emits widget updates.
 func TestSubagentMonitor_SubagentLifecycle(t *testing.T) {
 	harness := test.New(t)
-	harness.LoadFile("../../.kit/extensions/subagent-monitor.go")
+	harness.LoadFile("./subagent-monitor.go")

 	// Start session
 	_, err := harness.Emit(extensions.SessionStartEvent{SessionID: "test-session"})
@@ -84,7 +84,7 @@ func TestSubagentMonitor_SubagentLifecycle(t *testing.T) {
 // TestSubagentMonitor_MultipleSubagents verifies multiple parallel subagents.
 func TestSubagentMonitor_MultipleSubagents(t *testing.T) {
 	harness := test.New(t)
-	harness.LoadFile("../../.kit/extensions/subagent-monitor.go")
+	harness.LoadFile("./subagent-monitor.go")

 	_, err := harness.Emit(extensions.SessionStartEvent{SessionID: "test-session"})
 	if err != nil {
@@ -134,7 +134,7 @@ func TestSubagentMonitor_MultipleSubagents(t *testing.T) {
 // subagents emit events concurrently from different goroutines.
 func TestSubagentMonitor_ConcurrentSubagents(t *testing.T) {
 	harness := test.New(t)
-	harness.LoadFile("../../.kit/extensions/subagent-monitor.go")
+	harness.LoadFile("./subagent-monitor.go")

 	_, err := harness.Emit(extensions.SessionStartEvent{SessionID: "test-session"})
 	if err != nil {
@@ -186,7 +186,7 @@ func TestSubagentMonitor_ConcurrentSubagents(t *testing.T) {
 // even with nil ctx functions.
 func TestSubagentMonitor_SessionShutdown(t *testing.T) {
 	harness := test.New(t)
-	harness.LoadFile("../../.kit/extensions/subagent-monitor.go")
+	harness.LoadFile("./subagent-monitor.go")

 	// Start then shutdown
 	_, err := harness.Emit(extensions.SessionStartEvent{SessionID: "test-session"})
@@ -1,11 +1,11 @@
 module github.com/mark3labs/kit

-go 1.26.2
+go 1.26.3

 require (
 	charm.land/bubbles/v2 v2.1.0
 	charm.land/bubbletea/v2 v2.0.6
-	charm.land/fantasy v0.23.0
+	charm.land/fantasy v0.25.0
 	charm.land/huh/v2 v2.0.3
 	charm.land/lipgloss/v2 v2.0.3
 	github.com/alecthomas/chroma/v2 v2.24.1
@@ -14,19 +14,19 @@ require (
 	github.com/charmbracelet/fang v1.0.0
 	github.com/charmbracelet/log v1.0.0
 	github.com/charmbracelet/openai-go v0.0.0-20260319145158-d0740cc34266
-	github.com/charmbracelet/ultraviolet v0.0.0-20260428153724-66037269d7be
+	github.com/charmbracelet/ultraviolet v0.0.0-20260511121909-c840852527f3
 	github.com/charmbracelet/x/editor v0.2.0
 	github.com/clipperhouse/displaywidth v0.11.0
 	github.com/clipperhouse/uax29/v2 v2.7.0
-	github.com/coder/acp-go-sdk v0.12.2
+	github.com/coder/acp-go-sdk v0.13.0
 	github.com/fsnotify/fsnotify v1.10.1
 	github.com/indaco/herald v0.13.0
 	github.com/indaco/herald-md v0.3.0
-	github.com/mark3labs/mcp-go v0.51.0
+	github.com/mark3labs/mcp-go v0.54.0
 	github.com/spf13/cobra v1.10.2
 	github.com/spf13/viper v1.21.0
 	github.com/traefik/yaegi v0.16.1
-	golang.org/x/term v0.42.0
+	golang.org/x/term v0.43.0
 	gopkg.in/yaml.v3 v3.0.1
 )

@@ -59,9 +59,9 @@ require (
 	github.com/charmbracelet/harmonica v0.2.0 // indirect
 	github.com/charmbracelet/lipgloss v1.1.1-0.20250404203927-76690c660834 // indirect
 	github.com/charmbracelet/x/cellbuf v0.0.15 // indirect
-	github.com/charmbracelet/x/exp/charmtone v0.0.0-20260503005035-c113ba3d2310 // indirect
+	github.com/charmbracelet/x/exp/charmtone v0.0.0-20260519012233-798e623c8447 // indirect
 	github.com/charmbracelet/x/exp/ordered v0.1.0 // indirect
-	github.com/charmbracelet/x/exp/slice v0.0.0-20260503005035-c113ba3d2310 // indirect
+	github.com/charmbracelet/x/exp/slice v0.0.0-20260519012233-798e623c8447 // indirect
 	github.com/charmbracelet/x/exp/strings v0.1.0 // indirect
 	github.com/charmbracelet/x/json v0.2.0 // indirect
 	github.com/charmbracelet/x/termios v0.1.1 // indirect
@@ -69,7 +69,7 @@ require (
 	github.com/dlclark/regexp2 v1.12.0 // indirect
 	github.com/dustin/go-humanize v1.0.1 // indirect
 	github.com/felixge/httpsnoop v1.0.4 // indirect
-	github.com/go-json-experiment/json v0.0.0-20260430182902-b6187a392ed4 // indirect
+	github.com/go-json-experiment/json v0.0.0-20260505212615-e40f80bf6836 // indirect
 	github.com/go-logfmt/logfmt v0.6.1 // indirect
 	github.com/go-logr/logr v1.4.3 // indirect
 	github.com/go-logr/stdr v1.2.2 // indirect
@@ -82,10 +82,10 @@ require (
 	github.com/googleapis/enterprise-certificate-proxy v0.3.15 // indirect
 	github.com/googleapis/gax-go/v2 v2.22.0 // indirect
 	github.com/gorilla/websocket v1.5.3 // indirect
-	github.com/kaptinlin/go-i18n v0.4.7 // indirect
-	github.com/kaptinlin/jsonpointer v0.4.21 // indirect
-	github.com/kaptinlin/jsonschema v0.7.13 // indirect
-	github.com/kaptinlin/messageformat-go v0.6.3 // indirect
+	github.com/kaptinlin/go-i18n v0.4.9 // indirect
+	github.com/kaptinlin/jsonpointer v0.4.24 // indirect
+	github.com/kaptinlin/jsonschema v0.7.14 // indirect
+	github.com/kaptinlin/messageformat-go v0.6.4 // indirect
 	github.com/mitchellh/hashstructure/v2 v2.0.2 // indirect
 	github.com/muesli/mango v0.2.0 // indirect
 	github.com/muesli/mango-cobra v1.3.0 // indirect
@@ -97,7 +97,7 @@ require (
 	github.com/spf13/afero v1.15.0 // indirect
 	github.com/spf13/cast v1.10.0 // indirect
 	github.com/subosito/gotenv v1.6.0 // indirect
-	github.com/tidwall/gjson v1.18.0 // indirect
+	github.com/tidwall/gjson v1.19.0 // indirect
 	github.com/tidwall/match v1.2.0 // indirect
 	github.com/tidwall/pretty v1.2.1 // indirect
 	github.com/tidwall/sjson v1.2.5 // indirect
@@ -111,15 +111,15 @@ require (
 	go.opentelemetry.io/otel/metric v1.43.0 // indirect
 	go.opentelemetry.io/otel/trace v1.43.0 // indirect
 	go.yaml.in/yaml/v3 v3.0.4 // indirect
-	golang.org/x/crypto v0.50.0 // indirect
-	golang.org/x/exp v0.0.0-20260410095643-746e56fc9e2f // indirect
-	golang.org/x/net v0.53.0 // indirect
+	golang.org/x/crypto v0.51.0 // indirect
+	golang.org/x/exp v0.0.0-20260508232706-74f9aab9d74a // indirect
+	golang.org/x/net v0.54.0 // indirect
 	golang.org/x/oauth2 v0.36.0 // indirect
 	golang.org/x/time v0.15.0 // indirect
-	google.golang.org/api v0.277.0 // indirect
-	google.golang.org/genai v1.55.0 // indirect
-	google.golang.org/genproto/googleapis/rpc v0.0.0-20260427160629-7cedc36a6bc4 // indirect
-	google.golang.org/grpc v1.81.0 // indirect
+	google.golang.org/api v0.279.0 // indirect
+	google.golang.org/genai v1.57.0 // indirect
+	google.golang.org/genproto/googleapis/rpc v0.0.0-20260519071638-aa98bba5eb94 // indirect
+	google.golang.org/grpc v1.81.1 // indirect
 	google.golang.org/protobuf v1.36.11 // indirect
 	gopkg.in/yaml.v2 v2.4.0 // indirect
 )
@@ -137,6 +137,6 @@ require (
 	github.com/rivo/uniseg v0.4.7 // indirect
 	github.com/spf13/pflag v1.0.10
 	golang.org/x/sync v0.20.0 // indirect
-	golang.org/x/sys v0.43.0 // indirect
-	golang.org/x/text v0.36.0
+	golang.org/x/sys v0.44.0 // indirect
+	golang.org/x/text v0.37.0
 )
@@ -2,8 +2,8 @@ charm.land/bubbles/v2 v2.1.0 h1:YSnNh5cPYlYjPxRrzs5VEn3vwhtEn3jVGRBT3M7/I0g=
 charm.land/bubbles/v2 v2.1.0/go.mod h1:l97h4hym2hvWBVfmJDtrEHHCtkIKeTEb3TTJ4ZOB3wY=
 charm.land/bubbletea/v2 v2.0.6 h1:UHN/91OyuhaOFGSrBXQ/hMZD8IO1Uc4BvHlgHXL2WJo=
 charm.land/bubbletea/v2 v2.0.6/go.mod h1:MH/D8ZLlN3op37vQvijKuU29g3rqTp+aQapURFonF9g=
-charm.land/fantasy v0.23.0 h1:pocjwC5CxfEg1Bpwb0raML2d5ijo3op33Mmd6hYJyo4=
-charm.land/fantasy v0.23.0/go.mod h1:4yzSsd9XmFEVjRnF1P0LTEbLTmQX6OLnPkrHaf7iruo=
+charm.land/fantasy v0.25.0 h1:oXOWY1ivmTSnhYGzAolscF8zKtavWZyBWv0LHRSwN5Q=
+charm.land/fantasy v0.25.0/go.mod h1:8QrWUzIcKwZQP+aAnC9vLu3iID6hu9/Jt+rPMiieBkc=
 charm.land/huh/v2 v2.0.3 h1:2cJsMqEPwSywGHvdlKsJyQKPtSJLVnFKyFbsYZTlLkU=
 charm.land/huh/v2 v2.0.3/go.mod h1:93eEveeeqn47MwiC3tf+2atZ2l7Is88rAtmZNZ8x9Wc=
 charm.land/lipgloss/v2 v2.0.3 h1:yM2zJ4Cf5Y51b7RHIwioil4ApI/aypFXXVHSwlM6RzU=
@@ -86,8 +86,8 @@ github.com/charmbracelet/log v1.0.0 h1:HVVVMmfOorfj3BA9i8X8UL69Hoz9lI0PYwXfJvOdR
 github.com/charmbracelet/log v1.0.0/go.mod h1:uYgY3SmLpwJWxmlrPwXvzVYujxis1vAKRV/0VQB7yWA=
 github.com/charmbracelet/openai-go v0.0.0-20260319145158-d0740cc34266 h1:BW/sZtyd1JyYy0h5adMm3tzpNyL857LWjuTRET6OhpY=
 github.com/charmbracelet/openai-go v0.0.0-20260319145158-d0740cc34266/go.mod h1:1DahUaExbUZx/jD+FNT2PKP4L9rLE5+ZBRuI8mZjd/E=
-github.com/charmbracelet/ultraviolet v0.0.0-20260428153724-66037269d7be h1:j7w8VP/D4lu5+/4GamMmFy8nrtadcl82/fjvDgSHwLo=
-github.com/charmbracelet/ultraviolet v0.0.0-20260428153724-66037269d7be/go.mod h1:3YdTxlnV/L0bQ3VN8WOSw8doF7LZV/xawUQ4MuAPDvo=
+github.com/charmbracelet/ultraviolet v0.0.0-20260511121909-c840852527f3 h1:pxGjlWZFcRQMWAdtjRelpL3Gbu8iYIyuO3Eqbd037Ow=
+github.com/charmbracelet/ultraviolet v0.0.0-20260511121909-c840852527f3/go.mod h1:SnKWaPaTnkTNXJgdgdquu66de12V8pW/b/qlTGaF9xg=
 github.com/charmbracelet/x/ansi v0.11.7 h1:kzv1kJvjg2S3r9KHo8hDdHFQLEqn4RBCb39dAYC84jI=
 github.com/charmbracelet/x/ansi v0.11.7/go.mod h1:9qGpnAVYz+8ACONkZBUWPtL7lulP9No6p1epAihUZwQ=
 github.com/charmbracelet/x/cellbuf v0.0.15 h1:ur3pZy0o6z/R7EylET877CBxaiE1Sp1GMxoFPAIztPI=
@@ -98,14 +98,14 @@ github.com/charmbracelet/x/editor v0.2.0 h1:7XLUKtaRaB8jN7bWU2p2UChiySyaAuIfYiIR
 github.com/charmbracelet/x/editor v0.2.0/go.mod h1:p3oQ28TSL3YPd+GKJ1fHWcp+7bVGpedHpXmo0D6t1dY=
 github.com/charmbracelet/x/errors v0.0.0-20240508181413-e8d8b6e2de86 h1:JSt3B+U9iqk37QUU2Rvb6DSBYRLtWqFqfxf8l5hOZUA=
 github.com/charmbracelet/x/errors v0.0.0-20240508181413-e8d8b6e2de86/go.mod h1:2P0UgXMEa6TsToMSuFqKFQR+fZTO9CNGUNokkPatT/0=
-github.com/charmbracelet/x/exp/charmtone v0.0.0-20260503005035-c113ba3d2310 h1:rByFKh9JgQScu7oy0+TlUbC2e93woW/QNZmNXbbbw/E=
-github.com/charmbracelet/x/exp/charmtone v0.0.0-20260503005035-c113ba3d2310/go.mod h1:nsExn0DGyX0lh9LwLHTn2Gg+hafdzfSXnC+QmEJTZFY=
+github.com/charmbracelet/x/exp/charmtone v0.0.0-20260519012233-798e623c8447 h1:ZJxL6oPAQXzr21550ad/ZNkHKC3Ita70bl4xVZUPqqE=
+github.com/charmbracelet/x/exp/charmtone v0.0.0-20260519012233-798e623c8447/go.mod h1:nsExn0DGyX0lh9LwLHTn2Gg+hafdzfSXnC+QmEJTZFY=
 github.com/charmbracelet/x/exp/golden v0.0.0-20250806222409-83e3a29d542f h1:pk6gmGpCE7F3FcjaOEKYriCvpmIN4+6OS/RD0vm4uIA=
 github.com/charmbracelet/x/exp/golden v0.0.0-20250806222409-83e3a29d542f/go.mod h1:IfZAMTHB6XkZSeXUqriemErjAWCCzT0LwjKFYCZyw0I=
 github.com/charmbracelet/x/exp/ordered v0.1.0 h1:55/qLwjIh0gL0Vni+QAWk7T/qRVP6sBf+2agPBgnOFE=
 github.com/charmbracelet/x/exp/ordered v0.1.0/go.mod h1:5UHwmG+is5THxMyCJHNPCn2/ecI07aKNrW+LcResjJ8=
-github.com/charmbracelet/x/exp/slice v0.0.0-20260503005035-c113ba3d2310 h1:PMjHdSo8Vpq9psUw9BoHo9JLPMkm9Hqb+Whk64n3AQQ=
-github.com/charmbracelet/x/exp/slice v0.0.0-20260503005035-c113ba3d2310/go.mod h1:vqEfX6xzqW1pKKZUUiFOKg0OQ7bCh54Q2vR/tserrRA=
+github.com/charmbracelet/x/exp/slice v0.0.0-20260519012233-798e623c8447 h1:dZNZoFaaoQYXmtAOz4ovm0kISwcKX5Xt29ZLqhNYQKc=
+github.com/charmbracelet/x/exp/slice v0.0.0-20260519012233-798e623c8447/go.mod h1:vqEfX6xzqW1pKKZUUiFOKg0OQ7bCh54Q2vR/tserrRA=
 github.com/charmbracelet/x/exp/strings v0.1.0 h1:i69S2XI7uG1u4NLGeJPSYU++Nmjvpo9nwd6aoEm7gkA=
 github.com/charmbracelet/x/exp/strings v0.1.0/go.mod h1:/ehtMPNh9K4odGFkqYJKpIYyePhdp1hLBRvyY4bWkH8=
 github.com/charmbracelet/x/json v0.2.0 h1:DqB+ZGx2h+Z+1s98HOuOyli+i97wsFQIxP2ZQANTPrQ=
@@ -124,8 +124,8 @@ github.com/clipperhouse/uax29/v2 v2.7.0 h1:+gs4oBZ2gPfVrKPthwbMzWZDaAFPGYK72F0NJ
 github.com/clipperhouse/uax29/v2 v2.7.0/go.mod h1:EFJ2TJMRUaplDxHKj1qAEhCtQPW2tJSwu5BF98AuoVM=
 github.com/cncf/xds/go v0.0.0-20260202195803-dba9d589def2 h1:aBangftG7EVZoUb69Os8IaYg++6uMOdKK83QtkkvJik=
 github.com/cncf/xds/go v0.0.0-20260202195803-dba9d589def2/go.mod h1:qwXFYgsP6T7XnJtbKlf1HP8AjxZZyzxMmc+Lq5GjlU4=
-github.com/coder/acp-go-sdk v0.12.2 h1:fpRJ8Z5HMSr5cZ5IywzFlFZcIxZOsto+laNVu7XelFA=
-github.com/coder/acp-go-sdk v0.12.2/go.mod h1:yKzM/3R9uELp4+nBAwwtkS0aN1FOFjo11CNPy37yFko=
+github.com/coder/acp-go-sdk v0.13.0 h1:IAKBDIbe/iBfKAGikeIndzb8fowt4ioD+gCtSU4HwMA=
+github.com/coder/acp-go-sdk v0.13.0/go.mod h1:yKzM/3R9uELp4+nBAwwtkS0aN1FOFjo11CNPy37yFko=
 github.com/cpuguy83/go-md2man/v2 v2.0.6/go.mod h1:oOW0eioCTA6cOiMLiUPZOpcVxMig6NIQQ7OS05n1F4g=
 github.com/creack/pty v1.1.24 h1:bJrF4RRfyJnbTJqzRLHzcGaZK1NeM5kTC9jGgovnR1s=
 github.com/creack/pty v1.1.24/go.mod h1:08sCNb52WyoAwi2QDyzUCTgcvVFhUzewun7wtTfvcwE=
@@ -148,8 +148,8 @@ github.com/frankban/quicktest v1.14.6 h1:7Xjx+VpznH+oBnejlPUj8oUpdxnVs4f8XU8WnHk
 github.com/frankban/quicktest v1.14.6/go.mod h1:4ptaffx2x8+WTWXmUCuVU6aPUX1/Mz7zb5vbUoiM6w0=
 github.com/fsnotify/fsnotify v1.10.1 h1:b0/UzAf9yR5rhf3RPm9gf3ehBPpf0oZKIjtpKrx59Ho=
 github.com/fsnotify/fsnotify v1.10.1/go.mod h1:TLheqan6HD6GBK6PrDWyDPBaEV8LspOxvPSjC+bVfgo=
-github.com/go-json-experiment/json v0.0.0-20260430182902-b6187a392ed4 h1:2WmHkJINIjgXXYDGik8d3oJvFA3DAwPy00csDJ3vo+o=
-github.com/go-json-experiment/json v0.0.0-20260430182902-b6187a392ed4/go.mod h1:tphK2c80bpPhMOI4v6bIc2xWywPfbqi1Z06+RcrMkDg=
+github.com/go-json-experiment/json v0.0.0-20260505212615-e40f80bf6836 h1:5KGUhXZFTN1PrCY4zUZLe1J8n7uBNmPDbCLCn78EbPQ=
+github.com/go-json-experiment/json v0.0.0-20260505212615-e40f80bf6836/go.mod h1:tphK2c80bpPhMOI4v6bIc2xWywPfbqi1Z06+RcrMkDg=
 github.com/go-logfmt/logfmt v0.6.1 h1:4hvbpePJKnIzH1B+8OR/JPbTx37NktoI9LE2QZBBkvE=
 github.com/go-logfmt/logfmt v0.6.1/go.mod h1:EV2pOAQoZaT1ZXZbqDl5hrymndi4SY9ED9/z6CO0XAk=
 github.com/go-logr/logr v1.2.2/go.mod h1:jdQByPbusPIv2/zmleS9BjJVeZ6kBagPoEUsqbVz/1A=
@@ -187,14 +187,14 @@ github.com/indaco/herald v0.13.0 h1:+xVG9Fx5NpuWhwku/9IlRL6I009NnX4VUGKvlZHTRxU=
 github.com/indaco/herald v0.13.0/go.mod h1:T5g1+XLYvpjouhzAGHnAHDCKizhESkoV6+QPZ3DhgWA=
 github.com/indaco/herald-md v0.3.0 h1:hN1cKyrexPPM9PeHBsKuaWvIizSi/iYvM9yzRgtdb8M=
 github.com/indaco/herald-md v0.3.0/go.mod h1:RUHVaDSG45ymJjKyxpDwBocLXrZo93FB4OeYMsw9B9s=
-github.com/kaptinlin/go-i18n v0.4.7 h1:apjIIZHnGRyrkiX3vHj07F1BF6D0JLmV+VGSr1781Jc=
-github.com/kaptinlin/go-i18n v0.4.7/go.mod h1:+i1J0pFq/9i9ESC5qRMVkKwC+mdQTABhhBExpYOlbeM=
-github.com/kaptinlin/jsonpointer v0.4.21 h1:WVkwQbeerbHFcoXG7Yo/mlQhhZjWiTnagECEfwDXXa0=
-github.com/kaptinlin/jsonpointer v0.4.21/go.mod h1:Mo7+DX8RlQTFqS4dnYJl0izSP4ob+Rl5xO/mGDETgaU=
-github.com/kaptinlin/jsonschema v0.7.13 h1:kahVXTy/rURL0XJjyQ9WELm59wEmXi6IY0TWswQEFvU=
-github.com/kaptinlin/jsonschema v0.7.13/go.mod h1:Uh0aUBusnhXDCEXJ2oimL/hx7YTo7F+sKniE+tM0ERc=
-github.com/kaptinlin/messageformat-go v0.6.3 h1:m9ZE/fCjnsk8bdkv7Qs56L/ZoHbmQqhz9mRZSAQLU5g=
-github.com/kaptinlin/messageformat-go v0.6.3/go.mod h1:2KOZ/hgo/SveZ+uyi7vPUpUXieX65Mppzbc3VpGyqKs=
+github.com/kaptinlin/go-i18n v0.4.9 h1:kxYNfExb1GG241BIPnQaD3YOvWkP7poHwfj8ctwIpkw=
+github.com/kaptinlin/go-i18n v0.4.9/go.mod h1:15vHHYLhwo1stIdztJatujSttIY6sRFt+7v8K1ik1AA=
+github.com/kaptinlin/jsonpointer v0.4.24 h1:cHAiQXWzRYagJmtvURLmCdUM06rQy/3Lji+q6b58Fxo=
+github.com/kaptinlin/jsonpointer v0.4.24/go.mod h1:wVOBaXGGnP42YsMb6zev/3W5POTvspdNfh8DXzf8XS8=
+github.com/kaptinlin/jsonschema v0.7.14 h1:6grzaTJiRuLXlIGEdlGX5HEII3Za2tV+xxGpW3Kg4Rc=
+github.com/kaptinlin/jsonschema v0.7.14/go.mod h1:9WFuBzJjrvNkXVjo0L2Ujl1T/yqAGurwgbx4JWgF5C8=
+github.com/kaptinlin/messageformat-go v0.6.4 h1:6nC70fsqEn2xxg/Xoby2+Dk2r77kvxa3QNnYL/hsNcM=
+github.com/kaptinlin/messageformat-go v0.6.4/go.mod h1:553UGZ1x5jmGtyH4pQKYwLGMyPm71deCoZICjq1DtR8=
 github.com/kr/pretty v0.3.1 h1:flRD4NNwYAUpkphVc1HcthR4KEIFJ65n8Mw5qdRn3LE=
 github.com/kr/pretty v0.3.1/go.mod h1:hoEshYVHaxMs3cyo3Yncou5ZscifuDolrwPKZanG3xk=
 github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
@@ -203,8 +203,8 @@ github.com/kylelemons/godebug v1.1.0 h1:RPNrshWIDI6G2gRW9EHilWtl7Z6Sb1BR0xunSBf0
 github.com/kylelemons/godebug v1.1.0/go.mod h1:9/0rRGxNHcop5bhtWyNeEfOS8JIWk580+fNqagV/RAw=
 github.com/lucasb-eyer/go-colorful v1.4.0 h1:UtrWVfLdarDgc44HcS7pYloGHJUjHV/4FwW4TvVgFr4=
 github.com/lucasb-eyer/go-colorful v1.4.0/go.mod h1:R4dSotOR9KMtayYi1e77YzuveK+i7ruzyGqttikkLy0=
-github.com/mark3labs/mcp-go v0.51.0 h1:e8AhEfxzcYt7XqYzwT7uzWNhnqpu3H1Tn7dEJB9Ygj8=
-github.com/mark3labs/mcp-go v0.51.0/go.mod h1:Zg9cB2HdwdMMVgY0xtTzq3KvYIOJQDsaut+jWjwDaQY=
+github.com/mark3labs/mcp-go v0.54.0 h1:PZhQvd+5xrT43cUoiaKn/hDcvLUhcLc1twSEKYPTcTA=
+github.com/mark3labs/mcp-go v0.54.0/go.mod h1:+8WclSK1ZUweCP3hvktSji8n8ABG/95QaEkeVE/Uwas=
 github.com/mattn/go-isatty v0.0.22 h1:j8l17JJ9i6VGPUFUYoTUKPSgKe/83EYU2zBC7YNKMw4=
 github.com/mattn/go-isatty v0.0.22/go.mod h1:ZXfXG4SQHsB/w3ZeOYbR0PrPwLy+n6xiMrJlRFqopa4=
 github.com/mattn/go-runewidth v0.0.23 h1:7ykA0T0jkPpzSvMS5i9uoNn2Xy3R383f9HDx3RybWcw=
@@ -256,8 +256,8 @@ github.com/stretchr/testify v1.11.1/go.mod h1:wZwfW3scLgRK+23gO65QZefKpKQRnfz6sD
 github.com/subosito/gotenv v1.6.0 h1:9NlTDc1FTs4qu0DDq7AEtTPNw6SVm7uBMsUCUjABIf8=
 github.com/subosito/gotenv v1.6.0/go.mod h1:Dk4QP5c2W3ibzajGcXpNraDfq2IrhjMIvMSWPKKo0FU=
 github.com/tidwall/gjson v1.14.2/go.mod h1:/wbyibRr2FHMks5tjHJ5F8dMZh3AcwJEMf5vlfC0lxk=
-github.com/tidwall/gjson v1.18.0 h1:FIDeeyB800efLX89e5a8Y0BNH+LOngJyGrIWxG2FKQY=
-github.com/tidwall/gjson v1.18.0/go.mod h1:/wbyibRr2FHMks5tjHJ5F8dMZh3AcwJEMf5vlfC0lxk=
+github.com/tidwall/gjson v1.19.0 h1:xwxm7n691Uf3u5OFjzngavjGTh55KX5q/9w9xHW88JU=
+github.com/tidwall/gjson v1.19.0/go.mod h1:V37/opeE/JbLUOfH0QTXiNez2l0RUjYUhpT4szFQAfc=
 github.com/tidwall/match v1.1.1/go.mod h1:eRSPERbgtNPcGhD8UCthc6PmLEQXEWd3PRB5JTxsfmM=
 github.com/tidwall/match v1.2.0 h1:0pt8FlkOwjN2fPt4bIl4BoNxb98gGHN2ObFEDkrfZnM=
 github.com/tidwall/match v1.2.0/go.mod h1:eRSPERbgtNPcGhD8UCthc6PmLEQXEWd3PRB5JTxsfmM=
@@ -292,38 +292,38 @@ go.opentelemetry.io/otel/trace v1.43.0 h1:BkNrHpup+4k4w+ZZ86CZoHHEkohws8AY+WTX09
 go.opentelemetry.io/otel/trace v1.43.0/go.mod h1:/QJhyVBUUswCphDVxq+8mld+AvhXZLhe+8WVFxiFff0=
 go.yaml.in/yaml/v3 v3.0.4 h1:tfq32ie2Jv2UxXFdLJdh3jXuOzWiL1fo0bu/FbuKpbc=
 go.yaml.in/yaml/v3 v3.0.4/go.mod h1:DhzuOOF2ATzADvBadXxruRBLzYTpT36CKvDb3+aBEFg=
-golang.org/x/crypto v0.50.0 h1:zO47/JPrL6vsNkINmLoo/PH1gcxpls50DNogFvB5ZGI=
-golang.org/x/crypto v0.50.0/go.mod h1:3muZ7vA7PBCE6xgPX7nkzzjiUq87kRItoJQM1Yo8S+Q=
-golang.org/x/exp v0.0.0-20260410095643-746e56fc9e2f h1:W3F4c+6OLc6H2lb//N1q4WpJkhzJCK5J6kUi1NTVXfM=
-golang.org/x/exp v0.0.0-20260410095643-746e56fc9e2f/go.mod h1:J1xhfL/vlindoeF/aINzNzt2Bket5bjo9sdOYzOsU80=
-golang.org/x/net v0.53.0 h1:d+qAbo5L0orcWAr0a9JweQpjXF19LMXJE8Ey7hwOdUA=
-golang.org/x/net v0.53.0/go.mod h1:JvMuJH7rrdiCfbeHoo3fCQU24Lf5JJwT9W3sJFulfgs=
+golang.org/x/crypto v0.51.0 h1:IBPXwPfKxY7cWQZ38ZCIRPI50YLeevDLlLnyC5wRGTI=
+golang.org/x/crypto v0.51.0/go.mod h1:8AdwkbraGNABw2kOX6YFPs3WM22XqI4EXEd8g+x7Oc8=
+golang.org/x/exp v0.0.0-20260508232706-74f9aab9d74a h1:+3jdDGGB8NGb1Zktc737jlt3/A5f6UlwSzmvqUuufxw=
+golang.org/x/exp v0.0.0-20260508232706-74f9aab9d74a/go.mod h1:d2fgXJLVs4dYDHUk5lwMIfzRzSrWCfGZb0ZqeLa/Vcw=
+golang.org/x/net v0.54.0 h1:2zJIZAxAHV/OHCDTCOHAYehQzLfSXuf/5SoL/Dv6w/w=
+golang.org/x/net v0.54.0/go.mod h1:Sj4oj8jK6XmHpBZU/zWHw3BV3abl4Kvi+Ut7cQcY+cQ=
 golang.org/x/oauth2 v0.36.0 h1:peZ/1z27fi9hUOFCAZaHyrpWG5lwe0RJEEEeH0ThlIs=
 golang.org/x/oauth2 v0.36.0/go.mod h1:YDBUJMTkDnJS+A4BP4eZBjCqtokkg1hODuPjwiGPO7Q=
 golang.org/x/sync v0.20.0 h1:e0PTpb7pjO8GAtTs2dQ6jYa5BWYlMuX047Dco/pItO4=
 golang.org/x/sync v0.20.0/go.mod h1:9xrNwdLfx4jkKbNva9FpL6vEN7evnE43NNNJQ2LF3+0=
-golang.org/x/sys v0.43.0 h1:Rlag2XtaFTxp19wS8MXlJwTvoh8ArU6ezoyFsMyCTNI=
-golang.org/x/sys v0.43.0/go.mod h1:4GL1E5IUh+htKOUEOaiffhrAeqysfVGipDYzABqnCmw=
-golang.org/x/term v0.42.0 h1:UiKe+zDFmJobeJ5ggPwOshJIVt6/Ft0rcfrXZDLWAWY=
-golang.org/x/term v0.42.0/go.mod h1:Dq/D+snpsbazcBG5+F9Q1n2rXV8Ma+71xEjTRufARgY=
-golang.org/x/text v0.36.0 h1:JfKh3XmcRPqZPKevfXVpI1wXPTqbkE5f7JA92a55Yxg=
-golang.org/x/text v0.36.0/go.mod h1:NIdBknypM8iqVmPiuco0Dh6P5Jcdk8lJL0CUebqK164=
+golang.org/x/sys v0.44.0 h1:ildZl3J4uzeKP07r2F++Op7E9B29JRUy+a27EibtBTQ=
+golang.org/x/sys v0.44.0/go.mod h1:4GL1E5IUh+htKOUEOaiffhrAeqysfVGipDYzABqnCmw=
+golang.org/x/term v0.43.0 h1:S4RLU2sB31O/NCl+zFN9Aru9A/Cq2aqKpTZJ6B+DwT4=
+golang.org/x/term v0.43.0/go.mod h1:lrhlHNdQJHO+1qVYiHfFKVuVioJIheAc3fBSMFYEIsk=
+golang.org/x/text v0.37.0 h1:Cqjiwd9eSg8e0QAkyCaQTNHFIIzWtidPahFWR83rTrc=
+golang.org/x/text v0.37.0/go.mod h1:a5sjxXGs9hsn/AJVwuElvCAo9v8QYLzvavO5z2PiM38=
 golang.org/x/time v0.15.0 h1:bbrp8t3bGUeFOx08pvsMYRTCVSMk89u4tKbNOZbp88U=
 golang.org/x/time v0.15.0/go.mod h1:Y4YMaQmXwGQZoFaVFk4YpCt4FLQMYKZe9oeV/f4MSno=
 gonum.org/v1/gonum v0.17.0 h1:VbpOemQlsSMrYmn7T2OUvQ4dqxQXU+ouZFQsZOx50z4=
 gonum.org/v1/gonum v0.17.0/go.mod h1:El3tOrEuMpv2UdMrbNlKEh9vd86bmQ6vqIcDwxEOc1E=
-google.golang.org/api v0.277.0 h1:HJfyJUiNeBBUMai7ez8u14wkp/gH/I4wpGbbO9o+cSk=
-google.golang.org/api v0.277.0/go.mod h1:B9TqLBwJqVjp1mtt7WeoQwWRwvu/400y5lETOql+giQ=
-google.golang.org/genai v1.55.0 h1:iLHGk4Bj/IZ/GNNZb7hYqwSJMRBvqLeu2Hb6YQ+rYGw=
-google.golang.org/genai v1.55.0/go.mod h1:A3kkl0nyBjyFlNjgxIwKq70julKbIxpSxqKO5gw/gmk=
-google.golang.org/genproto v0.0.0-20260427160629-7cedc36a6bc4 h1:2iMJZntwvmfgtse+s744JY7v7PgEdSBuFYXucvpOHNM=
-google.golang.org/genproto v0.0.0-20260427160629-7cedc36a6bc4/go.mod h1:v14kaaboYyXQ1Gsu489Q+Hg/oN4B33mWtuOhF1HCeXA=
-google.golang.org/genproto/googleapis/api v0.0.0-20260427160629-7cedc36a6bc4 h1:yOzSCGPx+cp5VO7IxvZ9SBFF7j1tZVcNtlHR2iYKtVo=
-google.golang.org/genproto/googleapis/api v0.0.0-20260427160629-7cedc36a6bc4/go.mod h1:Q9HWtNeE7tM9npdIsEvqXj1QJIvVoeAV3rtXtS715Cw=
-google.golang.org/genproto/googleapis/rpc v0.0.0-20260427160629-7cedc36a6bc4 h1:tEkOQcXgF6dH1G+MVKZrfpYvozGrzb91k6ha7jireSM=
-google.golang.org/genproto/googleapis/rpc v0.0.0-20260427160629-7cedc36a6bc4/go.mod h1:4Hqkh8ycfw05ld/3BWL7rJOSfebL2Q+DVDeRgYgxUU8=
-google.golang.org/grpc v1.81.0 h1:W3G9N3KQf3BU+YuCtGKJk0CmxQNbAISICD/9AORxLIw=
-google.golang.org/grpc v1.81.0/go.mod h1:xGH9GfzOyMTGIOXBJmXt+BX/V0kcdQbdcuwQ/zNw42I=
+google.golang.org/api v0.279.0 h1:hsx2M2OaRcaKtVYK6vXEUnQvdjnend7ZYES+lYaot74=
+google.golang.org/api v0.279.0/go.mod h1:B9TqLBwJqVjp1mtt7WeoQwWRwvu/400y5lETOql+giQ=
+google.golang.org/genai v1.57.0 h1:qTyG2ynz5dQy2jF4CvZdLHHVslhR0heMue+zM1a4GNM=
+google.golang.org/genai v1.57.0/go.mod h1:A3kkl0nyBjyFlNjgxIwKq70julKbIxpSxqKO5gw/gmk=
+google.golang.org/genproto v0.0.0-20260504160031-60b97b32f348 h1:JjVGDZYWkJWZcxveJGzfkXC5myDVWAd4dZdgbzrDUv8=
+google.golang.org/genproto v0.0.0-20260504160031-60b97b32f348/go.mod h1:95PqD4xM+AdOcBGsmgfaofXsiA37uXDtDufVbntT3TU=
+google.golang.org/genproto/googleapis/api v0.0.0-20260504160031-60b97b32f348 h1:U8orV30l6KpDsi9dxU0CoJZGbjS8EEpw+6ba+XwGPQA=
+google.golang.org/genproto/googleapis/api v0.0.0-20260504160031-60b97b32f348/go.mod h1:Yzdzr5OOZFgSsEV2D/Xi9NL3bszpXFAg0hFJiRohcD8=
+google.golang.org/genproto/googleapis/rpc v0.0.0-20260519071638-aa98bba5eb94 h1:eZCjr/aAF8c5ccm5pb6T4EXgIei5MlAAPWPJk+5ArfY=
+google.golang.org/genproto/googleapis/rpc v0.0.0-20260519071638-aa98bba5eb94/go.mod h1:4Hqkh8ycfw05ld/3BWL7rJOSfebL2Q+DVDeRgYgxUU8=
+google.golang.org/grpc v1.81.1 h1:VnnIIZ88UzOOKLukQi+ImGz8O1Wdp8nAGGnvOfEIWQQ=
+google.golang.org/grpc v1.81.1/go.mod h1:xGH9GfzOyMTGIOXBJmXt+BX/V0kcdQbdcuwQ/zNw42I=
 google.golang.org/protobuf v1.36.11 h1:fV6ZwhNocDyBLK0dj+fg8ektcVegBBuEolpbTQyBNVE=
 google.golang.org/protobuf v1.36.11/go.mod h1:HTf+CrKn2C3g5S8VImy6tdcUvCska2kB7j23XfzDpco=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
@@ -6,6 +6,7 @@ import (
 	"fmt"
 	"io"
 	"strings"
+	"sync"
 	"time"

 	"charm.land/fantasy"
@@ -585,8 +586,13 @@ func (a *Agent) GenerateWithCallbacks(ctx context.Context, messages []fantasy.Me
 	// This avoids type conflicts with provider-level options.
 	history = applyCacheControlToMessages(history)

-	// Track current tool call args for callbacks
-	var currentToolArgs string
+	// Track tool call args per-ToolCallID so parallel tool calls in a single
+	// step don't clobber each other. Without this, OnToolResult callbacks would
+	// all see the args of the last OnToolCall in the step. The mutex guards
+	// against the possibility that the underlying streaming layer dispatches
+	// callbacks from multiple goroutines.
+	toolCallArgs := make(map[string]string)
+	var toolCallArgsMu sync.Mutex

 	// Use the streaming path when streaming is enabled OR when any callbacks are
 	// provided. The agent only exposes tool/step callbacks on AgentStreamCall, so
@@ -773,7 +779,9 @@ func (a *Agent) GenerateWithCallbacks(ctx context.Context, messages []fantasy.Me
 				if ctx.Err() != nil {
 					return ctx.Err()
 				}
-				currentToolArgs = tc.Input
+				toolCallArgsMu.Lock()
+				toolCallArgs[tc.ToolCallID] = tc.Input
+				toolCallArgsMu.Unlock()

 				// Notify about the tool call
 				if cb.OnToolCall != nil {
@@ -793,15 +801,22 @@ func (a *Agent) GenerateWithCallbacks(ctx context.Context, messages []fantasy.Me
 				if ctx.Err() != nil {
 					return ctx.Err()
 				}
+				// Look up the args recorded for this specific tool call. Delete
+				// the entry so the map doesn't accumulate across steps.
+				toolCallArgsMu.Lock()
+				args := toolCallArgs[tr.ToolCallID]
+				delete(toolCallArgs, tr.ToolCallID)
+				toolCallArgsMu.Unlock()
+
 				// Notify tool execution finished
 				if cb.OnToolExecution != nil {
-					cb.OnToolExecution(tr.ToolCallID, tr.ToolName, currentToolArgs, false)
+					cb.OnToolExecution(tr.ToolCallID, tr.ToolName, args, false)
 				}

 				if cb.OnToolResult != nil {
 					// Extract result text and error status
 					resultText, isError := extractToolResultText(tr)
-					cb.OnToolResult(tr.ToolCallID, tr.ToolName, currentToolArgs, resultText, tr.ClientMetadata, isError)
+					cb.OnToolResult(tr.ToolCallID, tr.ToolName, args, resultText, tr.ClientMetadata, isError)
 				}

 				return nil
@@ -0,0 +1,109 @@
+package agent
+
+import (
+	"context"
+	"sync"
+	"testing"
+
+	"charm.land/fantasy"
+)
+
+// fakeParallelAgent simulates a provider that emits two parallel tool_use
+// blocks in a single step. It invokes the streaming callbacks in the order:
+//
+//	OnToolCall(A) -> OnToolCall(B) -> OnToolResult(A) -> OnToolResult(B)
+//
+// Before the fix in #33 the agent-layer wrapper recorded a single
+// `currentToolArgs` variable that was clobbered by the second OnToolCall, so
+// both OnToolResult callbacks received B's args instead of their own.
+type fakeParallelAgent struct {
+	calls   []fantasy.ToolCallContent
+	results []fantasy.ToolResultContent
+}
+
+func (f *fakeParallelAgent) Generate(_ context.Context, _ fantasy.AgentCall) (*fantasy.AgentResult, error) {
+	return &fantasy.AgentResult{}, nil
+}
+
+func (f *fakeParallelAgent) Stream(_ context.Context, opts fantasy.AgentStreamCall) (*fantasy.AgentResult, error) {
+	for _, tc := range f.calls {
+		if opts.OnToolCall != nil {
+			if err := opts.OnToolCall(tc); err != nil {
+				return nil, err
+			}
+		}
+	}
+	for _, tr := range f.results {
+		if opts.OnToolResult != nil {
+			if err := opts.OnToolResult(tr); err != nil {
+				return nil, err
+			}
+		}
+	}
+	return &fantasy.AgentResult{}, nil
+}
+
+// TestGenerateWithCallbacks_ParallelToolArgs is the regression test for #33.
+// It drives the streaming-callback wiring inside GenerateWithCallbacks with a
+// fake fantasy.Agent that emits two parallel tool calls before either result.
+// Each OnToolResult must receive the args of its own tool call (matched by
+// ToolCallID), not the args of the last OnToolCall in the step.
+func TestGenerateWithCallbacks_ParallelToolArgs(t *testing.T) {
+	t.Parallel()
+
+	argsA := `{"name":"scheduled_jobs"}`
+	argsB := `{"name":"gmail_trigger"}`
+
+	fake := &fakeParallelAgent{
+		calls: []fantasy.ToolCallContent{
+			{ToolCallID: "kit-A", ToolName: "load_skill", Input: argsA},
+			{ToolCallID: "kit-B", ToolName: "load_skill", Input: argsB},
+		},
+		results: []fantasy.ToolResultContent{
+			{ToolCallID: "kit-A", ToolName: "load_skill", Result: fantasy.ToolResultOutputContentText{Text: "ok-A"}},
+			{ToolCallID: "kit-B", ToolName: "load_skill", Result: fantasy.ToolResultOutputContentText{Text: "ok-B"}},
+		},
+	}
+
+	a := &Agent{
+		fantasyAgent:     fake,
+		streamingEnabled: false, // exercise the "hasCallbacks" branch
+	}
+
+	var mu sync.Mutex
+	resultArgs := map[string]string{}
+	executionArgs := map[string]string{} // captured when running == false
+
+	cb := GenerateCallbacks{
+		OnToolExecution: func(id, _, args string, running bool) {
+			if running {
+				return
+			}
+			mu.Lock()
+			defer mu.Unlock()
+			executionArgs[id] = args
+		},
+		OnToolResult: func(id, _, args, _, _ string, _ bool) {
+			mu.Lock()
+			defer mu.Unlock()
+			resultArgs[id] = args
+		},
+	}
+
+	if _, err := a.GenerateWithCallbacks(context.Background(), nil, cb); err != nil {
+		t.Fatalf("GenerateWithCallbacks returned error: %v", err)
+	}
+
+	if got, want := resultArgs["kit-A"], argsA; got != want {
+		t.Errorf("OnToolResult for kit-A: args = %q, want %q", got, want)
+	}
+	if got, want := resultArgs["kit-B"], argsB; got != want {
+		t.Errorf("OnToolResult for kit-B: args = %q, want %q", got, want)
+	}
+	if got, want := executionArgs["kit-A"], argsA; got != want {
+		t.Errorf("OnToolExecution(finish) for kit-A: args = %q, want %q", got, want)
+	}
+	if got, want := executionArgs["kit-B"], argsB; got != want {
+		t.Errorf("OnToolExecution(finish) for kit-B: args = %q, want %q", got, want)
+	}
+}
@@ -9,12 +9,19 @@ import (
 	"github.com/mark3labs/kit/internal/tools"
 )

+// mcpExecutor is the subset of *tools.MCPToolManager that the adapter
+// actually uses. Extracted as an interface so the adapter is unit-testable
+// without constructing a full manager + connection pool.
+type mcpExecutor interface {
+	ExecuteTool(ctx context.Context, prefixedName, inputJSON string) (*tools.MCPToolResult, error)
+}
+
 // mcpAgentTool adapts an tools.MCPTool to the fantasy.AgentTool interface.
 // This keeps the fantasy dependency confined to the agent layer — the tools
 // package is a pure MCP client library with no LLM framework dependency.
 type mcpAgentTool struct {
 	tool            tools.MCPTool
-	manager         *tools.MCPToolManager
+	exec            mcpExecutor
 	providerOptions fantasy.ProviderOptions
 }

@@ -29,10 +36,26 @@ func (t *mcpAgentTool) Info() fantasy.ToolInfo {
 }

 // Run executes the MCP tool by delegating to the MCPToolManager.
+//
+// MCP-side failures (JSON-RPC protocol errors, transport failures, schema
+// validation rejections from the server) are surfaced to the model as soft
+// tool errors rather than escalated to a critical agent error. This matches
+// the contract that native Kit tools follow via kit.ErrorResult(...) and
+// lets the model self-correct (e.g. retry with a fixed argument shape) or
+// give up gracefully rather than aborting the turn mid-run.
+//
+// Context cancellation is the one exception: if the caller cancelled the
+// context the turn was aborted intentionally, so we propagate the ctx error
+// to let the agent loop unwind cleanly.
 func (t *mcpAgentTool) Run(ctx context.Context, call fantasy.ToolCall) (fantasy.ToolResponse, error) {
-	result, err := t.manager.ExecuteTool(ctx, t.tool.Name, call.Input)
+	result, err := t.exec.ExecuteTool(ctx, t.tool.Name, call.Input)
 	if err != nil {
-		return fantasy.ToolResponse{}, fmt.Errorf("mcp tool execution failed: %w", err)
+		if ctxErr := ctx.Err(); ctxErr != nil {
+			return fantasy.ToolResponse{}, ctxErr
+		}
+		return fantasy.NewTextErrorResponse(
+			fmt.Sprintf("MCP tool %q failed: %s", t.tool.Name, err.Error()),
+		), nil
 	}

 	if result.IsError {
@@ -57,8 +80,8 @@ func mcpToolsToAgentTools(mcpTools []tools.MCPTool, manager *tools.MCPToolManage
 	agentTools := make([]fantasy.AgentTool, len(mcpTools))
 	for i, t := range mcpTools {
 		agentTools[i] = &mcpAgentTool{
-			tool:    t,
-			manager: manager,
+			tool: t,
+			exec: manager,
 		}
 	}
 	return agentTools
@@ -0,0 +1,158 @@
+package agent
+
+import (
+	"context"
+	"errors"
+	"strings"
+	"testing"
+	"time"
+
+	"charm.land/fantasy"
+
+	"github.com/mark3labs/kit/internal/tools"
+)
+
+// stubExecutor lets each test script the (result, err) pair returned by
+// ExecuteTool. The adapter holds an mcpExecutor interface, so this is the
+// only seam the tests need.
+type stubExecutor struct {
+	result *tools.MCPToolResult
+	err    error
+	// called records the last invocation for assertion.
+	called bool
+	name   string
+	input  string
+}
+
+func (s *stubExecutor) ExecuteTool(_ context.Context, prefixedName, inputJSON string) (*tools.MCPToolResult, error) {
+	s.called = true
+	s.name = prefixedName
+	s.input = inputJSON
+	return s.result, s.err
+}
+
+func newMCPAgentTool(exec mcpExecutor, name string) *mcpAgentTool {
+	return &mcpAgentTool{
+		tool: tools.MCPTool{Name: name},
+		exec: exec,
+	}
+}
+
+// Manager-side Go errors (JSON-RPC protocol errors, transport failures,
+// schema validation rejections from the MCP server) must be surfaced to
+// the model as soft tool errors so the agent loop can keep going. Aborting
+// the turn would discard all prior tool results — see issue #N.
+func TestMCPAgentTool_RPCErrorBecomesSoftError(t *testing.T) {
+	exec := &stubExecutor{
+		err: errors.New("MCP error -32602: Invalid params: missing field \"task\""),
+	}
+	tool := newMCPAgentTool(exec, "pubmed__search")
+
+	resp, err := tool.Run(context.Background(), fantasy.ToolCall{
+		ID:    "call-1",
+		Name:  "pubmed__search",
+		Input: `{"query":"foo"}`,
+	})
+
+	if err != nil {
+		t.Fatalf("expected nil error (soft), got %v", err)
+	}
+	if !resp.IsError {
+		t.Fatalf("expected IsError=true, got false")
+	}
+	if !strings.Contains(resp.Content, "pubmed__search") {
+		t.Errorf("expected tool name in error content, got %q", resp.Content)
+	}
+	if !strings.Contains(resp.Content, "-32602") {
+		t.Errorf("expected underlying error text in content, got %q", resp.Content)
+	}
+}
+
+// Context cancellation is the one error that must remain critical: it
+// means the caller intentionally aborted, and the agent loop needs to
+// unwind cleanly rather than burning more steps.
+func TestMCPAgentTool_CtxCancelStaysCritical(t *testing.T) {
+	exec := &stubExecutor{
+		// Real managers typically return ctx.Err() (or a wrapper) when the
+		// context is cancelled mid-call.
+		err: context.Canceled,
+	}
+	tool := newMCPAgentTool(exec, "slow__tool")
+
+	ctx, cancel := context.WithCancel(context.Background())
+	cancel()
+
+	resp, err := tool.Run(ctx, fantasy.ToolCall{Name: "slow__tool"})
+
+	if !errors.Is(err, context.Canceled) {
+		t.Fatalf("expected context.Canceled, got %v", err)
+	}
+	if resp.IsError || resp.Content != "" {
+		t.Errorf("expected empty response on critical error, got IsError=%v Content=%q", resp.IsError, resp.Content)
+	}
+}
+
+// Deadline-exceeded behaves the same as cancellation: ctx.Err() is
+// non-nil, so the adapter must propagate the critical error rather than
+// converting the executor's error into a soft response.
+func TestMCPAgentTool_CtxDeadlineStaysCritical(t *testing.T) {
+	exec := &stubExecutor{err: context.DeadlineExceeded}
+	tool := newMCPAgentTool(exec, "slow__tool")
+
+	ctx, cancel := context.WithDeadline(context.Background(), time.Now().Add(-time.Second))
+	defer cancel()
+
+	resp, err := tool.Run(ctx, fantasy.ToolCall{Name: "slow__tool"})
+	if !errors.Is(err, context.DeadlineExceeded) {
+		t.Fatalf("expected context.DeadlineExceeded, got %v", err)
+	}
+	if resp.IsError || resp.Content != "" {
+		t.Errorf("expected empty response on critical error, got IsError=%v Content=%q", resp.IsError, resp.Content)
+	}
+}
+
+// Server-side soft errors (CallToolResult{ isError: true }) must continue
+// to flow through as soft errors — this was the existing behavior and
+// must not regress.
+func TestMCPAgentTool_ServerIsErrorRemainsSoftError(t *testing.T) {
+	exec := &stubExecutor{
+		result: &tools.MCPToolResult{
+			IsError: true,
+			Content: "search service is rate limited; try again in 30s",
+		},
+	}
+	tool := newMCPAgentTool(exec, "pubmed__search")
+
+	resp, err := tool.Run(context.Background(), fantasy.ToolCall{Name: "pubmed__search"})
+	if err != nil {
+		t.Fatalf("expected nil error, got %v", err)
+	}
+	if !resp.IsError {
+		t.Fatalf("expected IsError=true, got false")
+	}
+	if resp.Content != "search service is rate limited; try again in 30s" {
+		t.Errorf("expected pass-through content, got %q", resp.Content)
+	}
+}
+
+// Happy path: ordinary successful tool result is passed through unchanged.
+func TestMCPAgentTool_SuccessIsPassthrough(t *testing.T) {
+	exec := &stubExecutor{
+		result: &tools.MCPToolResult{
+			IsError: false,
+			Content: `{"hits":3}`,
+		},
+	}
+	tool := newMCPAgentTool(exec, "pubmed__search")
+
+	resp, err := tool.Run(context.Background(), fantasy.ToolCall{Name: "pubmed__search"})
+	if err != nil {
+		t.Fatalf("unexpected error: %v", err)
+	}
+	if resp.IsError {
+		t.Fatalf("expected IsError=false")
+	}
+	if resp.Content != `{"hits":3}` {
+		t.Errorf("expected pass-through content, got %q", resp.Content)
+	}
+}
@@ -70,14 +70,24 @@ type App struct {
 	rootCtx    context.Context
 	rootCancel context.CancelFunc

-	// widgetUpdatePending is set to true when a WidgetUpdateEvent has been
-	// sent to the TUI but not yet consumed by its event loop. While the flag
-	// is set, subsequent NotifyWidgetUpdate calls are coalesced (dropped) to
-	// prevent fast extension tickers from flooding the BubbleTea mailbox with
-	// redundant re-render triggers. The flag is cleared after a short debounce
-	// (~1 frame) so new updates are always let through once the TUI has had a
-	// chance to process the pending event.
-	widgetUpdatePending atomic.Bool
+	// widgetUpdatePending is set to true while a WidgetUpdateEvent burst is
+	// being coalesced. The leading edge fires immediately; subsequent calls
+	// within the debounce window set widgetUpdateTrailing so a final event
+	// is delivered with the latest runner state at the end of the window.
+	// Without the trailing send, a rapid SetWidget→RemoveWidget pair (e.g.
+	// SubagentEnd pushing a final frame then removing the widget) would let
+	// the second call get silently dropped, leaving the TUI's layout stuck
+	// on the pre-removal widget height — visible as empty rows below the
+	// status bar after the widget disappears.
+	widgetUpdatePending  atomic.Bool
+	widgetUpdateTrailing atomic.Bool
+
+	// steerDrainFn is the test seam used by releaseBusyAfterCompact to pull
+	// any steer messages that arrived during compaction. In production it is
+	// nil and the helper falls back to a.opts.Kit.DrainSteer(); tests that
+	// need to exercise the steer-drain path without standing up a full
+	// *kit.Kit can set this field directly to inject fake items.
+	steerDrainFn func() []queueItem
 }

 // New creates a new App with the provided options and pre-loaded messages.
@@ -356,6 +366,10 @@ func (a *App) AddContextMessage(text string) {
 // tea.Program. customInstructions is optional text appended to the summary
 // prompt (e.g. "Focus on the API design decisions").
 //
+// Any prompts queued via Run/RunWithFiles or steering messages injected via
+// Steer/SteerWithFiles while compaction is running are flushed automatically
+// once compaction completes (see releaseBusyAfterCompact).
+//
 // Satisfies ui.AppController.
 func (a *App) CompactConversation(customInstructions string) error {
 	a.mu.Lock()
@@ -377,11 +391,7 @@ func (a *App) CompactConversation(customInstructions string) error {

 	go func() {
 		defer a.wg.Done()
-		defer func() {
-			a.mu.Lock()
-			a.busy = false
-			a.mu.Unlock()
-		}()
+		defer a.releaseBusyAfterCompact()

 		// Subscribe to SDK events for streaming compaction summary to the TUI.
 		sendFn := func(msg tea.Msg) {
@@ -420,6 +430,9 @@ func (a *App) CompactConversation(customInstructions string) error {
 // CompactAsync is like CompactConversation but calls onComplete/onError
 // callbacks instead of sending TUI events. Used by the extension API's
 // ctx.Compact() which needs callback-based notification.
+//
+// Like CompactConversation, any prompts/steer messages received during
+// compaction are flushed automatically once compaction finishes.
 func (a *App) CompactAsync(customInstructions string, onComplete func(), onError func(string)) error {
 	a.mu.Lock()
 	if a.closed {
@@ -440,11 +453,7 @@ func (a *App) CompactAsync(customInstructions string, onComplete func(), onError

 	go func() {
 		defer a.wg.Done()
-		defer func() {
-			a.mu.Lock()
-			a.busy = false
-			a.mu.Unlock()
-		}()
+		defer a.releaseBusyAfterCompact()

 		// Subscribe to SDK events for streaming compaction summary to the TUI.
 		sendFn := func(msg tea.Msg) {
@@ -489,6 +498,81 @@ func (a *App) CompactAsync(customInstructions string, onComplete func(), onError
 	return nil
 }

+// releaseBusyAfterCompact is the deferred tail that runs at the end of every
+// compaction goroutine (success, error, or panic-after-recover paths). It
+// flips a.busy back to false, but before doing so it checks whether any
+// prompts piled up while compaction was running:
+//
+//   - Run/RunWithFiles append to a.queue when a.busy is set.
+//   - Steer/SteerWithFiles deposit messages into the SDK steer channel via
+//     Kit.InjectSteerWithFiles when a.busy is set.
+//
+// Without this hand-off the queue would sit idle until the user submits
+// another prompt — see issue #27. If we find anything pending we keep busy
+// set, splice the steer messages to the front of the queue, and start a
+// fresh drainQueue goroutine to deliver them as a single batched turn.
+func (a *App) releaseBusyAfterCompact() {
+	// Pull steer messages outside the app mutex; DrainSteer takes its own
+	// internal lock and we don't want to nest the two. The test seam
+	// (a.steerDrainFn) takes precedence so unit tests can inject fake
+	// steer items without a real *kit.Kit.
+	var steerItems []queueItem
+	switch {
+	case a.steerDrainFn != nil:
+		steerItems = a.steerDrainFn()
+	case a.opts.Kit != nil:
+		if leftover := a.opts.Kit.DrainSteer(); len(leftover) > 0 {
+			steerItems = make([]queueItem, len(leftover))
+			for i, sm := range leftover {
+				steerItems[i] = queueItem{Prompt: sm.Text, Files: sm.Files}
+			}
+		}
+	}
+
+	a.mu.Lock()
+	// If the app was closed while compaction was running, drop everything
+	// and just clear busy. Run/Steer would have rejected new items already
+	// after Close(), but this guards against in-flight items that slipped
+	// in just before closed was set.
+	if a.closed {
+		a.queue = a.queue[:0]
+		a.busy = false
+		a.mu.Unlock()
+		return
+	}
+
+	// Combine steer-channel items (front) with the in-memory queue (back).
+	// Steer messages are placed first so they retain their "act now"
+	// semantics relative to ordinary queued prompts that arrived later.
+	pending := append(steerItems, a.queue...)
+	a.queue = a.queue[:0]
+
+	if len(pending) == 0 {
+		a.busy = false
+		a.mu.Unlock()
+		return
+	}
+
+	// Hand off to drainQueue: it will pick up the first item directly and
+	// scoop the rest from a.queue on its first iteration.
+	first := pending[0]
+	if len(pending) > 1 {
+		a.queue = append(a.queue, pending[1:]...)
+	}
+	// Stay busy across the goroutine swap.
+	a.wg.Add(1)
+	a.mu.Unlock()
+
+	// Notify the UI that steer-channel messages were consumed so the
+	// steering badge can clear; ordinary queued prompts will be reflected
+	// by the QueueUpdatedEvent that drainQueue emits as it picks them up.
+	if len(steerItems) > 0 {
+		a.sendEvent(SteerConsumedEvent{})
+	}
+
+	go a.drainQueue(first)
+}
+
 // --------------------------------------------------------------------------
 // Non-interactive execution
 // --------------------------------------------------------------------------
@@ -1076,32 +1160,47 @@ func (a *App) NotifyModelChanged(provider, model string) {
 // extension widgets. Called from the extension context's SetWidget/RemoveWidget
 // closures. In non-interactive mode this is a no-op (widgets are TUI-only).
 //
-// Coalescing: if a WidgetUpdateEvent is already queued and not yet consumed
-// by the TUI event loop, additional calls within the same ~16 ms window are
-// dropped. This prevents fast extension tickers from flooding BubbleTea's
-// mailbox with redundant re-render triggers.
+// Coalescing (leading + trailing edge): the first call in an idle period
+// fires immediately for responsiveness. Subsequent calls within a ~16 ms
+// debounce window are batched into a single trailing event delivered at
+// the end of the window. The trailing send is essential for correctness:
+// extensions routinely make tight SetWidget→RemoveWidget pairs (e.g. on
+// SubagentEnd) and silently dropping the second call would leave the TUI's
+// layout stuck on stale widget dimensions until some other event happens
+// to trigger a re-render.
 func (a *App) NotifyWidgetUpdate() {
-	// Coalesce: only one pending update at a time.
 	if !a.widgetUpdatePending.CompareAndSwap(false, true) {
+		// A leading-edge event is already in flight — mark that the runner
+		// state has changed again so the trailing send below picks it up.
+		a.widgetUpdateTrailing.Store(true)
 		return
 	}
 	a.mu.Lock()
 	prog := a.program
 	a.mu.Unlock()
-	if prog != nil {
-		prog.Send(WidgetUpdateEvent{})
-		// Reset the pending flag after a short debounce so subsequent calls
-		// within the same render cycle are also coalesced, but new updates
-		// after the cycle are allowed through.
-		go func() {
-			time.Sleep(16 * time.Millisecond) // ~1 frame at 60 fps
-			a.widgetUpdatePending.Store(false)
-		}()
-	} else {
+	if prog == nil {
 		// No program registered (non-interactive mode); clear the flag so
 		// future calls are never permanently blocked.
 		a.widgetUpdatePending.Store(false)
+		return
 	}
+	prog.Send(WidgetUpdateEvent{})
+	go func() {
+		time.Sleep(16 * time.Millisecond) // ~1 frame at 60 fps
+		// If any extra calls came in during the debounce window, deliver
+		// one trailing event so the TUI sees the latest widget state. We
+		// swap-and-test instead of plain-load so concurrent calls after
+		// the trailing send still race correctly with the pending reset.
+		if a.widgetUpdateTrailing.Swap(false) {
+			a.mu.Lock()
+			p := a.program
+			a.mu.Unlock()
+			if p != nil {
+				p.Send(WidgetUpdateEvent{})
+			}
+		}
+		a.widgetUpdatePending.Store(false)
+	}()
 }

 // NotifyContentReload sends a ContentReloadEvent to the TUI so it refreshes
@@ -763,3 +763,209 @@ func TestFormatMaxTokensTruncatedMessage_NoKit(t *testing.T) {
 		}
 	}
 }
+
+// --------------------------------------------------------------------------
+// releaseBusyAfterCompact (issue #27)
+// --------------------------------------------------------------------------
+
+// TestReleaseBusyAfterCompact_flushesQueuedMessages is a regression test for
+// issue #27: messages queued via Run() while /compact is running used to sit
+// in a.queue indefinitely until the user typed another prompt. After the fix
+// the deferred releaseBusyAfterCompact tail picks up any pending items and
+// dispatches drainQueue automatically.
+//
+// We simulate the compaction completion path directly (bypassing the SDK)
+// by toggling busy=true, populating the queue exactly as Run() would have
+// during compaction, and then invoking releaseBusyAfterCompact.
+func TestReleaseBusyAfterCompact_flushesQueuedMessages(t *testing.T) {
+	stub := newStubWithFuncs(
+		func(ctx context.Context) (*kit.TurnResult, error) {
+			return turnResult("compacted then drained"), nil
+		},
+	)
+	app := newTestApp(stub)
+	defer app.Close()
+
+	// Simulate the state at the start of the compaction tail: busy is set
+	// and a couple of prompts have piled up in the queue while we were
+	// summarising. (Run() would have appended them and returned a queue
+	// length > 0 to the caller.)
+	app.mu.Lock()
+	app.busy = true
+	app.queue = append(app.queue,
+		queueItem{Prompt: "queued during compact #1"},
+		queueItem{Prompt: "queued during compact #2"},
+	)
+	app.mu.Unlock()
+
+	// Invoke the deferred tail directly. It should kick off drainQueue.
+	app.releaseBusyAfterCompact()
+
+	// drainQueue runs in a goroutine. Wait for the app to come back to idle.
+	ok := waitForCondition(2*time.Second, func() bool {
+		app.mu.Lock()
+		defer app.mu.Unlock()
+		return !app.busy
+	})
+	if !ok {
+		t.Fatal("app did not become idle after releaseBusyAfterCompact: queue not drained")
+	}
+
+	// Wait for any in-flight goroutine to finish before reading state.
+	app.wg.Wait()
+
+	if got := app.QueueLength(); got != 0 {
+		t.Fatalf("expected empty queue after drain, got %d", got)
+	}
+	if n := stub.callCount(); n == 0 {
+		t.Fatalf("expected stub PromptFunc to fire at least once after compact, got %d calls", n)
+	}
+}
+
+// TestReleaseBusyAfterCompact_idleWhenQueueEmpty verifies that with no
+// pending messages the helper just clears busy and does NOT spawn a
+// drainQueue goroutine (no spurious agent turn).
+func TestReleaseBusyAfterCompact_idleWhenQueueEmpty(t *testing.T) {
+	stub := newStub()
+	app := newTestApp(stub)
+	defer app.Close()
+
+	app.mu.Lock()
+	app.busy = true
+	app.mu.Unlock()
+
+	app.releaseBusyAfterCompact()
+
+	app.mu.Lock()
+	busy := app.busy
+	app.mu.Unlock()
+	if busy {
+		t.Fatal("expected busy=false after releaseBusyAfterCompact with empty queue")
+	}
+
+	// Give any rogue goroutine a moment to (incorrectly) call PromptFunc.
+	time.Sleep(50 * time.Millisecond)
+	if n := stub.callCount(); n != 0 {
+		t.Fatalf("expected 0 PromptFunc calls when queue empty, got %d", n)
+	}
+}
+
+// TestReleaseBusyAfterCompact_splicesSteerAheadOfQueue exercises the SDK
+// steer-drain branch of releaseBusyAfterCompact (issue #27 follow-up).
+//
+// Production wires a.opts.Kit.DrainSteer() to pull messages that arrived via
+// Steer/SteerWithFiles during compaction, but Options.Kit is *kit.Kit (a
+// concrete struct) so unit tests cannot stand up a real instance without a
+// full LLM backend. The test uses the unexported steerDrainFn seam to inject
+// fake steer items, then asserts that:
+//
+//   - Steer items are dispatched ahead of any prompts that piled up in
+//     a.queue (steer retains "act now" priority over ordinary queued
+//     prompts), and
+//   - the helper still hands off to drainQueue so the steer item actually
+//     fires (the previous behaviour left them stranded — see #27).
+func TestReleaseBusyAfterCompact_splicesSteerAheadOfQueue(t *testing.T) {
+	var pmu sync.Mutex
+	var firstPrompt string
+	stub := newStubWithFuncs(
+		func(ctx context.Context) (*kit.TurnResult, error) {
+			return turnResult("steer dispatched"), nil
+		},
+	)
+	// Wrap PromptFunc so we can capture the prompt text the stub receives
+	// (newStubWithFuncs's fns ignore prompt; we need it to verify ordering).
+	capturingPrompt := func(ctx context.Context, prompt string) (*kit.TurnResult, error) {
+		pmu.Lock()
+		if firstPrompt == "" {
+			firstPrompt = prompt
+		}
+		pmu.Unlock()
+		return stub.fn(ctx, prompt)
+	}
+	app := New(Options{PromptFunc: capturingPrompt}, nil)
+	defer app.Close()
+
+	// Inject fake steer items via the test seam. In production the same
+	// items would have been delivered through Kit.InjectSteerWithFiles
+	// during /compact and pulled by DrainSteer here.
+	app.steerDrainFn = func() []queueItem {
+		return []queueItem{
+			{Prompt: "steer-1"},
+			{Prompt: "steer-2"},
+		}
+	}
+
+	// Simulate the state at the end of compaction: busy is set and a couple
+	// of regular Run() prompts have piled up after the steer messages.
+	app.mu.Lock()
+	app.busy = true
+	app.queue = append(app.queue,
+		queueItem{Prompt: "queued-1"},
+		queueItem{Prompt: "queued-2"},
+	)
+	app.mu.Unlock()
+
+	app.releaseBusyAfterCompact()
+
+	// Wait for the dispatched batch to complete.
+	ok := waitForCondition(2*time.Second, func() bool {
+		app.mu.Lock()
+		defer app.mu.Unlock()
+		return !app.busy
+	})
+	if !ok {
+		t.Fatal("app did not become idle after steer-spliced releaseBusyAfterCompact")
+	}
+	app.wg.Wait()
+
+	// drainQueue picks up `first` directly and batches the rest. With
+	// PromptFunc set, executeBatch invokes us with items[0] only — that
+	// item must be the first steer message, proving steer items were
+	// spliced ahead of the previously queued prompts.
+	pmu.Lock()
+	got := firstPrompt
+	pmu.Unlock()
+	if got != "steer-1" {
+		t.Fatalf("expected first dispatched prompt to be steer item %q (steer items must come before queued prompts), got %q",
+			"steer-1", got)
+	}
+
+	// Queue should be fully drained and PromptFunc must have actually fired.
+	if n := app.QueueLength(); n != 0 {
+		t.Fatalf("expected empty queue after drain, got %d entries", n)
+	}
+	if n := stub.callCount(); n == 0 {
+		t.Fatal("expected stub PromptFunc to fire at least once after splice")
+	}
+}
+
+// TestReleaseBusyAfterCompact_dropsQueueWhenClosed verifies that if the app
+// was closed during compaction the helper discards any pending items rather
+// than spawning drainQueue against a torn-down App.
+func TestReleaseBusyAfterCompact_dropsQueueWhenClosed(t *testing.T) {
+	stub := newStub()
+	app := newTestApp(stub)
+
+	app.mu.Lock()
+	app.busy = true
+	app.queue = append(app.queue, queueItem{Prompt: "would have run"})
+	app.closed = true
+	app.mu.Unlock()
+
+	app.releaseBusyAfterCompact()
+
+	app.mu.Lock()
+	busy := app.busy
+	qLen := len(app.queue)
+	app.mu.Unlock()
+	if busy {
+		t.Fatal("expected busy=false even when closed")
+	}
+	if qLen != 0 {
+		t.Fatalf("expected queue cleared on closed app, got %d entries", qLen)
+	}
+	time.Sleep(20 * time.Millisecond)
+	if n := stub.callCount(); n != 0 {
+		t.Fatalf("expected 0 PromptFunc calls on closed app, got %d", n)
+	}
+}
@@ -36,15 +36,17 @@ type Diagnostic struct {
 }

 // LoadAll discovers and loads all prompt templates from standard locations
-// and any extra paths. Templates are loaded in order of precedence (lowest
-// to highest), with later templates overriding earlier ones of the same name.
+// and any extra paths. Templates are loaded in order of precedence (highest
+// to lowest); the first source to define a given name wins, later definitions
+// of the same name are dropped with a diagnostic.
 //
 // Discovery paths searched in order:
 //  1. Default templates (if IncludeDefaults)
-//  2. ~/.kit/prompts/ (global user templates)
-//  3. .kit/prompts/ (project-local templates)
-//  4. ConfigPaths (from configuration)
-//  5. ExtraPaths (explicit paths, highest precedence)
+//  2. ~/.kit/prompts/                          (legacy global)
+//  3. $XDG_CONFIG_HOME/kit/prompts/            (XDG global, default ~/.config/kit/prompts/)
+//  4. <cwd>/.kit/prompts/                      (project-local templates)
+//  5. ConfigPaths                              (from configuration)
+//  6. ExtraPaths                               (explicit paths, lowest precedence)
 func LoadAll(opts LoadOptions) ([]*PromptTemplate, []Diagnostic, error) {
 	if opts.Cwd == "" {
 		opts.Cwd, _ = os.Getwd()
@@ -88,13 +90,21 @@ func LoadAll(opts LoadOptions) ([]*PromptTemplate, []Diagnostic, error) {
 		addTemplates(defaults, "default")
 	}

-	// 2. Global user templates: ~/.kit/prompts/
-	globalDir := filepath.Join(opts.HomeDir, ".kit", "prompts")
-	if templates, err := LoadFromDir(globalDir); err == nil {
+	// 2. Legacy global user templates: ~/.kit/prompts/
+	legacyGlobalDir := filepath.Join(opts.HomeDir, ".kit", "prompts")
+	if templates, err := LoadFromDir(legacyGlobalDir); err == nil {
 		addTemplates(templates, "global")
 	}

-	// 3. Project-local templates: .kit/prompts/
+	// 3. XDG global user templates: $XDG_CONFIG_HOME/kit/prompts/
+	//    Default: ~/.config/kit/prompts/. Aligns with extensions and skills.
+	if xdgDir := GlobalDir(); xdgDir != "" && xdgDir != legacyGlobalDir {
+		if templates, err := LoadFromDir(xdgDir); err == nil {
+			addTemplates(templates, "global")
+		}
+	}
+
+	// 4. Project-local templates: .kit/prompts/
 	localDir := filepath.Join(opts.Cwd, ".kit", "prompts")
 	if templates, err := LoadFromDir(localDir); err == nil {
 		addTemplates(templates, "local")
@@ -186,3 +196,22 @@ func loadDefaultTemplates() []*PromptTemplate {
 	// For now, return an empty slice - users can define their own templates
 	return nil
 }
+
+// GlobalDir returns the XDG-aligned global prompts directory, respecting
+// $XDG_CONFIG_HOME. Defaults to ~/.config/kit/prompts/. Returns an empty
+// string if the user's home directory cannot be resolved.
+//
+// This is the canonical location for user-wide prompt templates and aligns
+// with the discovery paths used for extensions ($XDG_CONFIG_HOME/kit/extensions/)
+// and skills ($XDG_CONFIG_HOME/kit/skills/).
+func GlobalDir() string {
+	base := os.Getenv("XDG_CONFIG_HOME")
+	if base == "" {
+		home, err := os.UserHomeDir()
+		if err != nil {
+			return ""
+		}
+		base = filepath.Join(home, ".config")
+	}
+	return filepath.Join(base, "kit", "prompts")
+}
@@ -129,26 +129,35 @@ func TestCompactionWithNewMessagesAfterCompaction(t *testing.T) {
 	msg4 := message.Message{Role: message.RoleAssistant, Parts: []message.ContentPart{message.TextContent{Text: "Message 4 - after compaction"}}}
 	_, _ = tm.AppendMessage(msg4)

-	// BuildContext should return: [summary] + [M4 (new after compaction)] + [M3 (kept)]
+	// BuildContext should return: [summary] + [M3 (kept)] + [M4 (new after compaction)]
+	// Kept messages must appear BEFORE post-compaction messages so the LLM
+	// sees the conversation in chronological order. Otherwise the latest
+	// post-compaction user message would be followed by an older kept user
+	// message, breaking user/assistant alternation and causing the model to
+	// respond as if the post-compaction turn never happened.
 	messages, _, _ := tm.BuildContext()
 	if len(messages) != 3 {
-		t.Fatalf("expected 3 messages (summary + M4 + M3), got %d: %+v", len(messages), messages)
+		t.Fatalf("expected 3 messages (summary + M3 + M4), got %d: %+v", len(messages), messages)
 	}

-	// Verify order: summary, M4 (new), M3 (kept)
+	// Verify order: summary, M3 (kept), M4 (new)
 	if messages[0].Role != fantasy.MessageRoleSystem {
 		t.Errorf("first message should be summary, got %s", messages[0].Role)
 	}
-	if messages[1].Role != fantasy.MessageRoleAssistant {
-		t.Errorf("second message should be assistant (M4), got %s", messages[1].Role)
+	if messages[1].Role != fantasy.MessageRoleUser {
+		t.Errorf("second message should be user (M3 kept), got %s", messages[1].Role)
 	}
-	m4Text := messages[1].Content[0].(fantasy.TextPart).Text
+	m3Text := messages[1].Content[0].(fantasy.TextPart).Text
+	if m3Text != "Message 3 - kept" {
+		t.Errorf("unexpected M3 text: %s", m3Text)
+	}
+	if messages[2].Role != fantasy.MessageRoleAssistant {
+		t.Errorf("third message should be assistant (M4 post-compact), got %s", messages[2].Role)
+	}
+	m4Text := messages[2].Content[0].(fantasy.TextPart).Text
 	if m4Text != "Message 4 - after compaction" {
 		t.Errorf("unexpected M4 text: %s", m4Text)
 	}
-	if messages[2].Role != fantasy.MessageRoleUser {
-		t.Errorf("third message should be user (M3), got %s", messages[2].Role)
-	}

 	// Verify that M1 is NOT in the context
 	for i, msg := range messages {
@@ -6,8 +6,10 @@ import (
 	"fmt"
 	"os"
 	"path/filepath"
+	"runtime"
 	"sort"
 	"strings"
+	"sync"
 	"time"
 )

@@ -97,6 +99,11 @@ func ListAllSessions() ([]SessionInfo, error) {

 // listSessionsInDir reads all .jsonl files in a directory and extracts session info.
 // Empty sessions (no messages) are automatically cleaned up and not returned.
+//
+// Per-file extraction is parallelized across a small worker pool because each
+// file requires a full JSONL scan to compute MessageCount and FirstMessage —
+// for users with many sessions this is the dominant cost of opening the
+// session picker.
 func listSessionsInDir(dir string) ([]SessionInfo, error) {
 	if _, err := os.Stat(dir); os.IsNotExist(err) {
 		return nil, nil
@@ -107,20 +114,47 @@ func listSessionsInDir(dir string) ([]SessionInfo, error) {
 		return nil, fmt.Errorf("failed to read directory %s: %w", dir, err)
 	}

-	var sessions []SessionInfo
+	// Collect candidate paths first so we can parallelize the heavy work.
+	paths := make([]string, 0, len(entries))
 	for _, entry := range entries {
 		if entry.IsDir() || !strings.HasSuffix(entry.Name(), ".jsonl") {
 			continue
 		}
+		paths = append(paths, filepath.Join(dir, entry.Name()))
+	}

-		path := filepath.Join(dir, entry.Name())
-		info, err := extractSessionInfo(path)
-		if err != nil {
-			continue // skip malformed session files
+	results := make([]*SessionInfo, len(paths))
+
+	// Worker pool sized to GOMAXPROCS, capped to avoid thrashing for tiny lists.
+	workers := max(min(runtime.GOMAXPROCS(0), len(paths)), 1)
+
+	var wg sync.WaitGroup
+	jobs := make(chan int, len(paths))
+	for range workers {
+		wg.Go(func() {
+			for i := range jobs {
+				info, err := extractSessionInfo(paths[i])
+				if err != nil {
+					continue // skip malformed session files
+				}
+				results[i] = info
+			}
+		})
+	}
+	for i := range paths {
+		jobs <- i
+	}
+	close(jobs)
+	wg.Wait()
+
+	sessions := make([]SessionInfo, 0, len(results))
+	for i, info := range results {
+		if info == nil {
+			continue
 		}
-		// Clean up and skip empty sessions (no messages)
+		// Clean up and skip empty sessions (no messages).
 		if info.MessageCount == 0 {
-			_ = os.Remove(path)
+			_ = os.Remove(paths[i])
 			continue
 		}
 		sessions = append(sessions, *info)
@@ -755,9 +755,17 @@ func (tm *TreeManager) BuildContext() (messages []fantasy.Message, provider stri
 		}
 	}

-	// If there is a compaction, inject the summary first and collect
-	// the kept messages starting from FirstKeptEntryID (since the
-	// compaction entry's parent chain doesn't include them).
+	// If there is a compaction, inject the summary first, then the
+	// preserved "kept" messages (chronologically before the compaction),
+	// then the post-compaction messages (chronologically after).
+	//
+	// Order matters: the kept messages must come BEFORE the post-compaction
+	// branch so the LLM sees the conversation in chronological order. If the
+	// kept messages were appended last, the latest user message in the
+	// current branch would be followed by an older kept user message,
+	// breaking the strict user/assistant alternation that providers expect
+	// and causing the model to respond as if the previous turn never
+	// happened.
 	if lastCompaction != nil {
 		messages = append(messages, fantasy.Message{
 			Role: fantasy.MessageRoleSystem,
@@ -768,49 +776,10 @@ func (tm *TreeManager) BuildContext() (messages []fantasy.Message, provider stri
 			},
 		})

-		// Collect entries from the compaction entry itself (at compactionIndex)
-		// and any entries before it in the branch (newer messages).
-		for i := compactionIndex; i < len(branch); i++ {
-			entry := branch[i]
-			switch e := entry.(type) {
-			case *MessageEntry:
-				msg, err := e.ToMessage()
-				if err != nil {
-					continue // skip malformed entries
-				}
-				msgs := msg.ToLLMMessages()
-				messages = append(messages, msgs...)
-
-			case *BranchSummaryEntry:
-				// Convert branch summary to a user message for context.
-				if e.Summary != "" {
-					messages = append(messages, fantasy.Message{
-						Role: fantasy.MessageRoleUser,
-						Content: []fantasy.MessagePart{
-							fantasy.TextPart{
-								Text: fmt.Sprintf("[Branch context: %s]", e.Summary),
-							},
-						},
-					})
-				}
-
-			case *ModelChangeEntry:
-				provider = e.Provider
-				modelID = e.ModelID
-
-			case *CompactionEntry:
-				// Already handled above (summary injected).
-				continue
-			}
-		}
-
-		// Now collect the kept messages starting from FirstKeptEntryID.
-		// These are not in the current branch because the compaction entry
-		// is parented to the first kept entry's parent, not the first kept entry.
-		// We iterate through entries in order (not using getBranchLocked) to avoid
-		// walking back to old compacted messages.
-		// We stop when we reach the compaction entry to avoid double-counting
-		// messages that were added after the compaction.
+		// Step 1: collect the kept messages starting from FirstKeptEntryID.
+		// These are not on the current branch (the compaction entry is a
+		// new root with no parent), so we iterate tm.entries in append order
+		// and stop when we reach the compaction entry itself.
 		if lastCompaction.FirstKeptEntryID != "" {
 			found := false
 			for _, entry := range tm.entries {
@@ -825,13 +794,12 @@ func (tm *TreeManager) BuildContext() (messages []fantasy.Message, provider stri
 					}
 				}

-				// Stop when we reach the compaction entry itself.
-				// Messages after the compaction are collected from the branch walk above.
+				// Stop when we reach the compaction entry itself; messages
+				// after it are collected from the branch walk below.
 				if entryID == lastCompaction.ID {
 					break
 				}

-				// Process this kept entry.
 				switch e := entry.(type) {
 				case *MessageEntry:
 					msg, err := e.ToMessage()
@@ -860,6 +828,42 @@ func (tm *TreeManager) BuildContext() (messages []fantasy.Message, provider stri
 			}
 		}

+		// Step 2: collect entries on the current branch after the compaction
+		// entry (these are post-compaction messages). The compaction entry
+		// itself is skipped — its summary was already injected above.
+		for i := compactionIndex; i < len(branch); i++ {
+			entry := branch[i]
+			switch e := entry.(type) {
+			case *MessageEntry:
+				msg, err := e.ToMessage()
+				if err != nil {
+					continue
+				}
+				msgs := msg.ToLLMMessages()
+				messages = append(messages, msgs...)
+
+			case *BranchSummaryEntry:
+				if e.Summary != "" {
+					messages = append(messages, fantasy.Message{
+						Role: fantasy.MessageRoleUser,
+						Content: []fantasy.MessagePart{
+							fantasy.TextPart{
+								Text: fmt.Sprintf("[Branch context: %s]", e.Summary),
+							},
+						},
+					})
+				}
+
+			case *ModelChangeEntry:
+				provider = e.Provider
+				modelID = e.ModelID
+
+			case *CompactionEntry:
+				// Summary already injected above.
+				continue
+			}
+		}
+
 		return messages, provider, modelID
 	}

@@ -1030,44 +1034,22 @@ func (tm *TreeManager) GetContextEntryIDs() []string {

 	var ids []string

-	// If there's a compaction, we need to collect IDs from:
-	// 1. Entries after the compaction entry in the branch (newer messages)
-	// 2. Entries from FirstKeptEntryID onwards (kept messages)
+	// If there's a compaction, we collect IDs in the same order as
+	// BuildContext: [summary placeholder, kept messages, post-compaction
+	// messages]. This ordering must stay in sync with BuildContext so a
+	// cut-point index can be mapped back to the correct entry ID.
 	if lastCompaction != nil {
 		// Placeholder for the summary system message (no entry ID).
 		ids = append(ids, "")

-		// Collect IDs from entries after the compaction entry (newer messages).
-		for i := compactionIndex + 1; i < len(branch); i++ {
-			entry := branch[i]
-			switch e := entry.(type) {
-			case *MessageEntry:
-				msg, err := e.ToMessage()
-				if err != nil {
-					continue
-				}
-				msgs := msg.ToLLMMessages()
-				for range msgs {
-					ids = append(ids, e.ID)
-				}
-
-			case *BranchSummaryEntry:
-				if e.Summary != "" {
-					ids = append(ids, e.ID)
-				}
-			}
-		}
-
-		// Collect IDs from the kept messages starting at FirstKeptEntryID.
-		// We iterate through entries in order (not using getBranchLocked) to avoid
-		// walking back to old compacted messages.
-		// We stop when we reach the compaction entry to avoid double-counting.
+		// Step 1: IDs of the kept messages starting at FirstKeptEntryID.
+		// Iterate tm.entries in append order and stop at the compaction
+		// entry to avoid double-counting post-compaction messages.
 		if lastCompaction.FirstKeptEntryID != "" {
 			found := false
 			for _, entry := range tm.entries {
 				entryID := tm.EntryID(entry)

-				// Skip entries until we reach the first kept entry.
 				if !found {
 					if entryID == lastCompaction.FirstKeptEntryID {
 						found = true
@@ -1076,7 +1058,6 @@ func (tm *TreeManager) GetContextEntryIDs() []string {
 					}
 				}

-				// Stop when we reach the compaction entry itself.
 				if entryID == lastCompaction.ID {
 					break
 				}
@@ -1100,6 +1081,28 @@ func (tm *TreeManager) GetContextEntryIDs() []string {
 			}
 		}

+		// Step 2: IDs of entries after the compaction entry on the current
+		// branch (post-compaction messages).
+		for i := compactionIndex + 1; i < len(branch); i++ {
+			entry := branch[i]
+			switch e := entry.(type) {
+			case *MessageEntry:
+				msg, err := e.ToMessage()
+				if err != nil {
+					continue
+				}
+				msgs := msg.ToLLMMessages()
+				for range msgs {
+					ids = append(ids, e.ID)
+				}
+
+			case *BranchSummaryEntry:
+				if e.Summary != "" {
+					ids = append(ids, e.ID)
+				}
+			}
+		}
+
 		return ids
 	}

@@ -161,6 +161,12 @@ var SlashCommands = []SlashCommand{
 		Category:    "Navigation",
 		Aliases:     []string{"/r"},
 	},
+	{
+		Name:        "/copy",
+		Description: "Copy the last message to the system clipboard",
+		Category:    "System",
+		Aliases:     []string{"/cp"},
+	},
 	{
 		Name:        "/export",
 		Description: "Export session (JSONL by default, or /export path.jsonl)",
@@ -129,8 +129,18 @@ type AppController interface {
 // SkillItem holds display metadata about a loaded skill for the startup
 // [Skills] section. Built by the CLI layer from the SDK's []*kit.Skill.
 type SkillItem struct {
-	Name   string // Skill name (e.g. "btca-cli").
-	Path   string // Absolute path to the skill file.
+	Name        string // Skill name (e.g. "btca-cli").
+	Path        string // Absolute path to the skill file.
+	Source      string // "project" or "user" (global).
+	Description string // Short summary used in autocomplete and help.
+}
+
+// ExtensionItem holds display metadata about a loaded extension for the
+// startup [Extensions] section. Built by the CLI layer from the SDK's
+// []kit.ExtensionInfo.
+type ExtensionItem struct {
+	Name   string // Extension display name (filename without .go extension).
+	Path   string // Absolute path to the extension's .go file.
 	Source string // "project" or "user" (global).
 }

@@ -363,6 +373,16 @@ type AppModelOptions struct {
 	// watcher detects changes. May be nil if skill hot-reload is not needed.
 	GetSkillItems func() []SkillItem

+	// ExtensionItems lists loaded extensions for the [Extensions] startup
+	// section. Each entry shows the filename of an extension that was
+	// discovered and loaded (global, project-local, or explicit).
+	ExtensionItems []ExtensionItem
+
+	// GetExtensionItems, if non-nil, returns the current extension items.
+	// Called on extension hot-reload to refresh the list. May be nil if no
+	// extensions are loaded.
+	GetExtensionItems func() []ExtensionItem
+
 	// MCPToolCount is the number of tools loaded from external MCP servers.
 	MCPToolCount int

@@ -607,6 +627,14 @@ type AppModel struct {
 	// skill list after content hot-reload. May be nil.
 	getSkillItems func() []SkillItem

+	// extensionItems lists loaded extensions for the [Extensions] startup
+	// section (filenames only).
+	extensionItems []ExtensionItem
+
+	// getExtensionItems returns the current extension items. Used to refresh
+	// the list after extension hot-reload. May be nil.
+	getExtensionItems func() []ExtensionItem
+
 	// mcpToolCount and extensionToolCount track tool counts by source for
 	// the startup info display.
 	mcpToolCount       int
@@ -860,6 +888,8 @@ func NewAppModel(appCtrl AppController, opts AppModelOptions) *AppModel {
 	m.contextPaths = opts.ContextPaths
 	m.skillItems = opts.SkillItems
 	m.getSkillItems = opts.GetSkillItems
+	m.extensionItems = opts.ExtensionItems
+	m.getExtensionItems = opts.GetExtensionItems
 	m.mcpToolCount = opts.MCPToolCount
 	m.extensionToolCount = opts.ExtensionToolCount
 	m.startupExtensionMessages = opts.StartupExtensionMessages
@@ -912,6 +942,20 @@ func NewAppModel(appCtrl AppController, opts AppModelOptions) *AppModel {
 		}
 	}

+	// Merge skills into autocomplete as /skill:<name> commands. Skills accept
+	// optional trailing args, so HasArgs is true — Enter populates the input
+	// with "/skill:name " rather than auto-submitting.
+	if ic, ok := m.input.(*InputComponent); ok && len(opts.SkillItems) > 0 {
+		for _, s := range opts.SkillItems {
+			ic.commands = append(ic.commands, commands.SlashCommand{
+				Name:        "/skill:" + s.Name,
+				Description: formatSkillDescription(s),
+				Category:    "Skills",
+				HasArgs:     true,
+			})
+		}
+	}
+
 	// Merge MCP prompts into autocomplete as /<server>:<prompt> commands.
 	if ic, ok := m.input.(*InputComponent); ok && len(opts.MCPPrompts) > 0 {
 		for _, p := range opts.MCPPrompts {
@@ -1014,8 +1058,21 @@ func (m *AppModel) AddStartupMessageToScrollList() {
 		pairs = append(pairs, [2]string{"Skills", strings.Join(names, ", ")})
 	}

-	// Extension tool count (only shown when > 0).
-	if m.extensionToolCount > 0 {
+	// Extensions — listed by filename. Each extension shows its basename
+	// without the .go suffix, matching the [Skills] section's style.
+	if len(m.extensionItems) > 0 {
+		names := make([]string, len(m.extensionItems))
+		for i, ei := range m.extensionItems {
+			names[i] = ei.Name
+		}
+		value := strings.Join(names, ", ")
+		if m.extensionToolCount > 0 {
+			value += fmt.Sprintf(" (%d tools)", m.extensionToolCount)
+		}
+		pairs = append(pairs, [2]string{"Extensions", value})
+	} else if m.extensionToolCount > 0 {
+		// Fallback: tool count only (extensions registered tools but the CLI
+		// did not provide ExtensionItems for some reason).
 		pairs = append(pairs, [2]string{"Extensions", fmt.Sprintf("%d tools", m.extensionToolCount)})
 	}

@@ -1251,7 +1308,11 @@ func (m *AppModel) Update(msg tea.Msg) (tea.Model, tea.Cmd) {
 			m.scrollList.autoScroll = false
 		case tea.MouseWheelDown:
 			m.scrollList.ScrollBy(scrollLines)
-			if m.scrollList.AtBottom() {
+			// Only re-enable auto-scroll when the user is not actively
+			// selecting text. Otherwise a wheel-down during a drag-select
+			// would re-arm GotoBottom on the next stream chunk, shifting
+			// the highlighted row out from under the cursor.
+			if m.scrollList.AtBottom() && !m.scrollList.IsMouseDown() {
 				m.scrollList.autoScroll = true
 			}
 		}
@@ -1259,9 +1320,14 @@ func (m *AppModel) Update(msg tea.Msg) (tea.Model, tea.Cmd) {
 	// ── Mouse click selection (crush-style character-level) ──────────────────
 	case tea.MouseClickMsg:
 		if msg.Button == tea.MouseLeft {
-			// Calculate viewport-relative coordinates.
-			viewY := msg.Y - m.scrollbackYOffset
-			if viewY >= 0 && viewY < m.scrollList.height {
+			// Compute the scrollback origin from the current frame's layout
+			// rather than the stale cached value from the previous View().
+			// scrollbackYOffset/scrollList.height are only refreshed inside
+			// View() and lag behind any state change that resized the header
+			// (extension widgets, warning rows, etc.) since the last render.
+			yOff, vpHeight := m.currentScrollbackBounds()
+			viewY := msg.Y - yOff
+			if viewY >= 0 && viewY < vpHeight {
 				// Clear any previous selection on a new click.
 				// HandleMouseDown will set up new selection state.
 				if m.scrollList.HandleMouseDown(msg.X, viewY) {
@@ -1272,8 +1338,9 @@ func (m *AppModel) Update(msg tea.Msg) (tea.Model, tea.Cmd) {

 	// ── Mouse motion/drag for character-level selection ──────────────────────
 	case tea.MouseMotionMsg:
-		viewY := msg.Y - m.scrollbackYOffset
-		if viewY >= 0 && viewY < m.scrollList.height {
+		yOff, vpHeight := m.currentScrollbackBounds()
+		viewY := msg.Y - yOff
+		if viewY >= 0 && viewY < vpHeight {
 			m.scrollList.HandleMouseDrag(msg.X, viewY)
 		}

@@ -1603,10 +1670,16 @@ func (m *AppModel) Update(msg tea.Msg) (tea.Model, tea.Cmd) {

 	// ── Cancel timer expired ─────────────────────────────────────────────────
 	case uicore.CancelTimerExpiredMsg:
+		if m.canceling {
+			m.layoutDirty = true
+		}
 		m.canceling = false

 	// ── Ctrl+C reset timer expired ────────────────────────────────────────────
 	case uicore.CtrlCResetMsg:
+		if m.ctrlCPressedOnce {
+			m.layoutDirty = true
+		}
 		m.ctrlCPressedOnce = false

 	// ── Input submitted ──────────────────────────────────────────────────────
@@ -2328,6 +2401,7 @@ func (m *AppModel) Update(msg tea.Msg) (tea.Model, tea.Cmd) {
 		if msg.err != nil {
 			m.printSystemMessage(fmt.Sprintf("Extension reload failed: %v", msg.err))
 		} else {
+			m.refreshExtensionItems()
 			m.printSystemMessage("Extensions reloaded.")
 		}

@@ -3095,6 +3169,8 @@ func (m *AppModel) handleSlashCommand(sc *commands.SlashCommand, args string) te
 		return m.handleResumeCommand()
 	case "/export":
 		return m.handleExportCommand(args)
+	case "/copy":
+		return m.handleCopyCommand()
 	case "/share":
 		return m.handleShareCommand()
 	case "/import":
@@ -3395,13 +3471,56 @@ func (m *AppModel) refreshPromptTemplates() {
 	}
 }

-// refreshSkillItems reloads skill items from the provider callback.
-// Called on ContentReloadEvent.
+// refreshSkillItems reloads skill items from the provider callback and
+// updates the autocomplete entries. Called on ContentReloadEvent.
 func (m *AppModel) refreshSkillItems() {
 	if m.getSkillItems == nil {
 		return
 	}
-	m.skillItems = m.getSkillItems()
+	newItems := m.getSkillItems()
+	m.skillItems = newItems
+
+	if ic, ok := m.input.(*InputComponent); ok {
+		// Remove old Skills commands and add fresh ones.
+		var kept []commands.SlashCommand
+		for _, sc := range ic.commands {
+			if sc.Category != "Skills" {
+				kept = append(kept, sc)
+			}
+		}
+		for _, s := range newItems {
+			kept = append(kept, commands.SlashCommand{
+				Name:        "/skill:" + s.Name,
+				Description: formatSkillDescription(s),
+				Category:    "Skills",
+				HasArgs:     true,
+			})
+		}
+		ic.commands = kept
+	}
+}
+
+// refreshExtensionItems reloads extension items from the provider callback
+// so the [Extensions] startup section reflects the current set after a
+// hot-reload. Called from the extReloadResultMsg handler.
+func (m *AppModel) refreshExtensionItems() {
+	if m.getExtensionItems == nil {
+		return
+	}
+	m.extensionItems = m.getExtensionItems()
+}
+
+// formatSkillDescription returns the autocomplete description for a skill,
+// prefixed with [project] or [user] so users can tell colliding names apart.
+func formatSkillDescription(s SkillItem) string {
+	prefix := "[user]"
+	if s.Source == "project" {
+		prefix = "[project]"
+	}
+	if s.Description == "" {
+		return prefix
+	}
+	return prefix + " " + s.Description
 }

 // refreshMCPPrompts reloads MCP prompts from the provider callback and
@@ -3476,6 +3595,7 @@ func (m *AppModel) printHelpMessage() {
 		"**System:**\n" +
 		"- `/compact [instructions]`: Summarise older messages to free context space\n" +
 		"- `/clear`: Clear message history\n" +
+		"- `/copy`: Copy the last message to the system clipboard\n" +
 		"- `/export [path]`: Export session as JSONL\n" +
 		"- `/import <path.jsonl>`: Import session from JSONL file\n" +
 		"- `/reset-usage`: Reset usage statistics\n" +
@@ -3712,7 +3832,12 @@ func (m *AppModel) appendStreamingChunk(role, content string) {
 		}
 		// Auto-scroll to bottom if enabled (iteratr pattern)
 		// Don't call SetItems() - the slice reference hasn't changed
-		if m.scrollList != nil {
+		//
+		// CRITICAL: never scroll the viewport while the user is actively
+		// selecting text (mouse button held). Doing so shifts the
+		// highlighted content out from under the cursor and produces the
+		// off-by-N-row drift users see when copy-selecting during streaming.
+		if m.scrollList != nil && !m.scrollList.IsMouseDown() {
 			if m.scrollList.autoScroll {
 				m.scrollList.GotoBottom()
 			} else if m.scrollList.AtBottom() {
@@ -3740,6 +3865,36 @@ func (m *AppModel) appendStreamingChunk(role, content string) {
 	m.refreshContent()
 }

+// currentScrollbackBounds returns the live (yOffset, viewportHeight) for the
+// scrollback region, computed from the current state — not from the cached
+// values populated inside View().
+//
+// scrollbackYOffset and scrollList.height are refreshed once per render, so
+// any state change that resizes the header (extension widget toggles,
+// warning rows, queued messages, etc.) leaves the cached values one frame
+// stale. Mouse click handlers in Update() can then place the cursor on the
+// wrong line, producing the off-by-N-row drift seen during copy-selection.
+//
+// This recomputes the header height by rendering it (cheap — the renderer
+// returns "" when no extension header is set) and recomputes the viewport
+// height the same way distributeHeight() does, so both inputs to the
+// y → (item, line) mapping are always current.
+func (m *AppModel) currentScrollbackBounds() (yOffset, viewportHeight int) {
+	// Force a fresh layout if anything in Update() marked the state dirty;
+	// otherwise scrollList.height still reflects the previous frame.
+	if m.layoutDirty {
+		m.distributeHeight()
+		m.layoutDirty = false
+	}
+	if headerView := m.renderHeaderFooter(m.getHeader); headerView != "" {
+		yOffset = lipgloss.Height(headerView)
+	}
+	if m.scrollList != nil {
+		viewportHeight = m.scrollList.height
+	}
+	return yOffset, viewportHeight
+}
+
 // distributeHeight recalculates child component heights after a window resize,
 // queue change, widget update, or state transition, and propagates the computed
 // stream height to the StreamComponent.
@@ -3812,7 +3967,20 @@ func (m *AppModel) distributeHeight() {
 		headerFooterLines += lipgloss.Height(footerView)
 	}

-	streamHeight := max(m.height-separatorLines-widgetLines-headerFooterLines-queuedLines-inputLines-statusBarLines, 0)
+	// Account for transient warning rows that View() injects between the
+	// scrollback and the separator. These flags are toggled by ESC/Ctrl+C
+	// handlers; without subtracting them here the joined view exceeds
+	// m.height by one line per active warning and the bottom of the screen
+	// gets silently clipped — which in turn invalidates scrollbackYOffset.
+	var warningLines int
+	if m.canceling {
+		warningLines++
+	}
+	if m.ctrlCPressedOnce {
+		warningLines++
+	}
+
+	streamHeight := max(m.height-separatorLines-widgetLines-headerFooterLines-queuedLines-inputLines-statusBarLines-warningLines, 0)

 	// In alt screen mode, give the calculated height to ScrollList instead of stream.
 	// The stream component still exists but is embedded as the last item in scrollList.
@@ -4236,6 +4404,48 @@ func (m *AppModel) handleNameCommand(args string) tea.Cmd {
 	return nil
 }

+// handleCopyCommand copies the last user or assistant message to the system
+// clipboard. Skips transient system messages (e.g. /help output) so the user
+// gets the actual last conversational message.
+func (m *AppModel) handleCopyCommand() tea.Cmd {
+	if len(m.messages) == 0 {
+		m.printSystemMessage("No messages to copy.")
+		return nil
+	}
+
+	var (
+		text string
+		role string
+	)
+	for i := len(m.messages) - 1; i >= 0; i-- {
+		switch msg := m.messages[i].(type) {
+		case *TextMessageItem:
+			if msg.role == "user" || msg.role == "assistant" {
+				text = msg.content
+				role = msg.role
+			}
+		case *StreamingMessageItem:
+			if msg.role == "assistant" || msg.role == "reasoning" {
+				text = msg.content.String()
+				role = msg.role
+			}
+		}
+		if text != "" {
+			break
+		}
+	}
+
+	if strings.TrimSpace(text) == "" {
+		m.printSystemMessage("No copyable message found.")
+		return nil
+	}
+
+	m.printSystemMessage(fmt.Sprintf(
+		"Copied last %s message to clipboard (%d chars).", role, len(text),
+	))
+	return clipboard.CopyToClipboard(text)
+}
+
 // handleExportCommand exports the current session to a file.
 // Usage: /export          — copies the JSONL file to cwd with a descriptive name.
 //
@@ -60,10 +60,13 @@ func NewScrollList(width, height int) *ScrollList {
 }

 // SetItems replaces the items in the scroll list. If auto-scroll is enabled,
-// the viewport will scroll to the bottom to show the latest content.
+// the viewport will scroll to the bottom to show the latest content — EXCEPT
+// when the user is actively selecting text (mouse button held), in which case
+// the scroll position is locked so the highlighted content stays under the
+// cursor. The pending bottom-scroll is deferred to MouseUp.
 func (s *ScrollList) SetItems(items []MessageItem) {
 	s.items = items
-	if s.autoScroll {
+	if s.autoScroll && !s.sel.MouseDown {
 		s.GotoBottom()
 	}
 }
@@ -157,6 +160,10 @@ func (s *ScrollList) HandleMouseDown(x, y int) bool {
 // HandleMouseDrag handles mouse motion while button is held.
 // Updates the selection endpoint for character-level precision.
 // Returns true if selection was updated.
+//
+// Defensively disables auto-scroll on every drag update — even if the
+// MouseDown handler missed (e.g. click landed in viewport padding), any
+// active drag means the user is selecting and the viewport must not jump.
 func (s *ScrollList) HandleMouseDrag(x, y int) bool {
 	if !s.sel.MouseDown {
 		return false
@@ -171,6 +178,9 @@ func (s *ScrollList) HandleMouseDrag(x, y int) bool {
 		return false
 	}

+	// Hard-lock the viewport while dragging.
+	s.autoScroll = false
+
 	s.sel.DragItemIdx = itemIdx
 	s.sel.DragLineIdx = lineIdx
 	s.sel.DragCol = x
@@ -178,6 +188,13 @@ func (s *ScrollList) HandleMouseDrag(x, y int) bool {
 	return true
 }

+// IsMouseDown reports whether the user currently has the mouse button held
+// (i.e. a selection drag is in progress). Used by the parent model to avoid
+// re-enabling auto-scroll during streaming while the user is selecting.
+func (s *ScrollList) IsMouseDown() bool {
+	return s.sel.MouseDown
+}
+
 // HandleMouseUp handles mouse button release.
 // Returns true if there was an active selection.
 func (s *ScrollList) HandleMouseUp() bool {
@@ -521,6 +538,21 @@ func (s *ScrollList) View() string {
 		for idx := s.offsetIdx; idx < len(s.items) && remainingHeight > 0; idx++ {
 			item := s.items[idx]
 			content := item.Render(s.width)
+
+			// Items that render to an empty string contribute zero height to
+			// the viewport. This MUST match renderedHeight()'s semantics —
+			// otherwise getItemAndLineAtY (which uses renderedHeight) treats
+			// the item as 0 lines while View() emits one blank line via
+			// strings.Split("", "\n") = [""], producing a 1-row downward
+			// drift in mouse hit-testing per empty item between offsetIdx
+			// and the cursor (most visibly streaming-reasoning items before
+			// any reasoning has streamed, which extension widgets surface by
+			// shrinking the scrollback).
+			if content == "" {
+				s.heightCache[item.ID()] = 0
+				continue
+			}
+
 			contentLines := strings.Split(content, "\n")

 			// Refresh height cache from the actual render (authoritative).
@@ -0,0 +1,181 @@
+package ui
+
+import (
+	"fmt"
+	"strings"
+	"testing"
+)
+
+// fakeItem is a deterministic MessageItem for ScrollList tests.
+type fakeItem struct {
+	id    string
+	lines int
+}
+
+func (f *fakeItem) ID() string { return f.id }
+func (f *fakeItem) Render(_ int) string {
+	if f.lines <= 0 {
+		return ""
+	}
+	parts := make([]string, f.lines)
+	for i := range parts {
+		parts[i] = fmt.Sprintf("%s-line-%d", f.id, i)
+	}
+	return strings.Join(parts, "\n")
+}
+func (f *fakeItem) Height() int { return f.lines }
+
+// makeItems builds n fake items of `lines` height each.
+func makeItems(n, lines int) []MessageItem {
+	out := make([]MessageItem, n)
+	for i := range out {
+		out[i] = &fakeItem{id: fmt.Sprintf("item-%d", i), lines: lines}
+	}
+	return out
+}
+
+// TestScrollList_MouseDownPreventsAutoScroll verifies the core fix for the
+// copy-selection drift bug: while the user has the mouse button held
+// (drag-selecting), incoming content updates must NOT shift the viewport,
+// because doing so moves the highlighted content out from under the cursor.
+func TestScrollList_MouseDownPreventsAutoScroll(t *testing.T) {
+	sl := NewScrollList(80, 10)
+	sl.SetItems(makeItems(20, 2)) // 40 lines of content into a 10-line viewport
+	// Capture the auto-scrolled-to-bottom position.
+	startOffsetIdx := sl.offsetIdx
+	startOffsetLine := sl.offsetLine
+
+	// User clicks somewhere in the visible area, starting a drag-select.
+	if !sl.HandleMouseDown(5, 3) {
+		t.Fatalf("HandleMouseDown should accept a click inside the viewport")
+	}
+	if !sl.IsMouseDown() {
+		t.Fatalf("IsMouseDown should be true after HandleMouseDown")
+	}
+
+	// New content arrives. With autoScroll still true, SetItems would
+	// normally call GotoBottom() and shift the viewport. The fix should
+	// suppress that while MouseDown is held.
+	sl.SetItems(makeItems(30, 2)) // 60 lines now
+	if sl.offsetIdx != startOffsetIdx || sl.offsetLine != startOffsetLine {
+		t.Errorf("viewport scrolled during active drag: was (%d,%d), now (%d,%d)",
+			startOffsetIdx, startOffsetLine, sl.offsetIdx, sl.offsetLine)
+	}
+
+	// User releases the mouse — drag is over.
+	sl.HandleMouseUp()
+	if sl.IsMouseDown() {
+		t.Fatalf("IsMouseDown should be false after HandleMouseUp")
+	}
+
+	// After release, a fresh content update should resume auto-scrolling
+	// (move the offset to track the new bottom).
+	afterReleaseIdx := sl.offsetIdx
+	afterReleaseLine := sl.offsetLine
+	sl.SetItems(makeItems(50, 2))
+	if sl.offsetIdx == afterReleaseIdx && sl.offsetLine == afterReleaseLine {
+		t.Errorf("autoscroll did not resume after MouseUp: offset stuck at (%d,%d)",
+			afterReleaseIdx, afterReleaseLine)
+	}
+}
+
+// TestScrollList_DragDisablesAutoScroll verifies that any successful
+// HandleMouseDrag call clears autoScroll, even when HandleMouseDown didn't
+// observe it (e.g. a stale wheel-down event set it back to true mid-stream).
+func TestScrollList_DragDisablesAutoScroll(t *testing.T) {
+	sl := NewScrollList(80, 10)
+	sl.SetItems(makeItems(20, 2))
+
+	// Begin a selection.
+	if !sl.HandleMouseDown(5, 3) {
+		t.Fatalf("HandleMouseDown failed")
+	}
+	// Simulate an external code path that re-enabled autoScroll while
+	// MouseDown is still held (the precise condition that caused drift).
+	sl.autoScroll = true
+
+	// Drag motion should hard-lock the viewport again.
+	if !sl.HandleMouseDrag(10, 4) {
+		t.Fatalf("HandleMouseDrag failed")
+	}
+	if sl.autoScroll {
+		t.Errorf("HandleMouseDrag must clear autoScroll to prevent mid-drag jumps")
+	}
+}
+
+// TestScrollList_SetItemsRespectsMouseDown is the most direct regression
+// test: even with autoScroll enabled and new content appended at the
+// bottom, SetItems must not move the viewport while a mouse drag is in
+// progress. This is what caused the "highlighting shifts by 1+ rows
+// during streaming" symptom reported by the user.
+func TestScrollList_SetItemsRespectsMouseDown(t *testing.T) {
+	sl := NewScrollList(80, 5)
+	sl.SetItems(makeItems(10, 2)) // 20 lines into a 5-line viewport
+	// At bottom.
+	preIdx, preLine := sl.offsetIdx, sl.offsetLine
+
+	// Hold mouse down (no actual drag needed).
+	if !sl.HandleMouseDown(0, 0) {
+		t.Fatalf("HandleMouseDown failed")
+	}
+
+	// Append several more items as if streaming. With the bug, each
+	// SetItems would call GotoBottom and shift the offset.
+	for n := 11; n <= 15; n++ {
+		sl.SetItems(makeItems(n, 2))
+		if sl.offsetIdx != preIdx || sl.offsetLine != preLine {
+			t.Fatalf("viewport drifted during streaming with mouse held: "+
+				"start=(%d,%d) now=(%d,%d) after adding item %d",
+				preIdx, preLine, sl.offsetIdx, sl.offsetLine, n)
+		}
+	}
+}
+
+// TestScrollList_EmptyItemsDoNotShiftMouseMapping is the regression test
+// for the second drift bug: items that render to "" must contribute the
+// same number of rows in View() (zero) as in renderedHeight(), or mouse
+// hit-testing drifts by one row per empty item between offsetIdx and the
+// cursor. This was surfaced by extension widgets (e.g. subagent-monitor)
+// that shrink the scrollback so empty streaming-reasoning items end up
+// in the visible window.
+//
+// Setup: 1 normal item + 1 empty item + 1 normal item. Click on the line
+// where the third item begins. With the bug, getItemAndLineAtY skips the
+// empty item (renderedHeight=0) and reports lineIdx pointing one row
+// past where View() actually painted that line.
+func TestScrollList_EmptyItemsDoNotShiftMouseMapping(t *testing.T) {
+	sl := NewScrollList(80, 10)
+	sl.SetItems([]MessageItem{
+		&fakeItem{id: "a", lines: 2},     // viewY 0–1
+		&fakeItem{id: "empty", lines: 0}, // renders "" — contributes 0 rows
+		&fakeItem{id: "b", lines: 2},     // viewY 2–3
+	})
+
+	// Render the viewport once so the cache reflects what View() actually
+	// emits (this is the path that previously diverged from renderedHeight
+	// for empty items).
+	rendered := sl.View()
+	lines := strings.Split(rendered, "\n")
+
+	// Sanity: View() must emit exactly height lines.
+	if len(lines) != 10 {
+		t.Fatalf("View() returned %d lines, want 10", len(lines))
+	}
+	// Item b's first line should appear at viewY=2, NOT viewY=3.
+	if !strings.Contains(lines[2], "b-line-0") {
+		t.Errorf("viewY=2 should render b-line-0 (empty item contributes 0 rows), got %q", lines[2])
+	}
+
+	// Now the actual hit-test contract: clicking on viewY=2 must map to
+	// item b line 0 — the same coordinate View() rendered there.
+	idx, line := sl.getItemAndLineAtY(2)
+	if idx != 2 || line != 0 {
+		t.Errorf("getItemAndLineAtY(2) = (%d,%d), want (2,0)", idx, line)
+	}
+
+	// And clicking on the second line of b (viewY=3) must map to b line 1.
+	idx, line = sl.getItemAndLineAtY(3)
+	if idx != 2 || line != 1 {
+		t.Errorf("getItemAndLineAtY(3) = (%d,%d), want (2,1)", idx, line)
+	}
+}
@@ -230,8 +230,10 @@ func FindWordBoundaries(line string, col int) (startCol, endCol int) {

 // HighlightLine applies reverse-video highlighting to a portion of a rendered
 // line (which may contain ANSI escape codes). startCol/endCol are in display
-// columns. If startCol == -1, the entire line is highlighted. If startCol ==
-// endCol, returns the line unchanged.
+// columns. If startCol == -1, the entire line is highlighted. If endCol ==
+// -1, the highlight runs from startCol to the end of the line (the sentinel
+// returned by IsLineInRange for the first line of a multi-line selection).
+// If startCol == endCol, returns the line unchanged.
 //
 // Uses ultraviolet ScreenBuffer for cell-level ANSI manipulation.
 func HighlightLine(line string, startCol, endCol int) string {
@@ -250,6 +252,16 @@ func HighlightLine(line string, startCol, endCol int) string {
 		endCol = lineWidth
 	}

+	// "From startCol to end of line" sentinel (returned by IsLineInRange
+	// for the first line of a multi-line selection). Without this branch,
+	// the start line of a multi-line drag would never be highlighted —
+	// the user perceives this as the selection being shifted one row down
+	// from the cursor, especially when extension widgets shrink the
+	// scrollback and make the start line land on a tall styled block.
+	if endCol < 0 {
+		endCol = lineWidth
+	}
+
 	if startCol >= endCol || startCol >= lineWidth {
 		return line
 	}
@@ -296,6 +308,11 @@ func ExtractText(line string, startCol, endCol int) string {
 		endCol = lineWidth
 	}

+	// "From startCol to end of line" sentinel (see HighlightLine).
+	if endCol < 0 {
+		endCol = lineWidth
+	}
+
 	if startCol >= endCol || startCol >= lineWidth {
 		return ""
 	}
@@ -357,6 +357,54 @@ func TestHighlightLine_NoSelection(t *testing.T) {
 	}
 }

+// TestHighlightLine_EndOfLineSentinel verifies that endCol=-1 is interpreted
+// as "highlight from startCol to end of line", matching the sentinel
+// returned by IsLineInRange for the first line of a multi-line selection.
+//
+// Regression: without this contract, the start line of any multi-line drag
+// would silently fall through HighlightLine's startCol >= endCol guard and
+// render unstyled, making the selection appear to begin one row below the
+// cursor — the exact "tracking gets shifted" symptom users reported when
+// extension widgets shrank the scrollback enough that the click landed on a
+// styled tool-result block.
+func TestHighlightLine_EndOfLineSentinel(t *testing.T) {
+	line := "Hello, World!"
+	result := HighlightLine(line, 0, -1)
+	if result == line {
+		t.Errorf("endCol=-1 should highlight from startCol to end of line; got unchanged input")
+	}
+	if len(result) <= len(line) {
+		t.Errorf("highlighted result should be longer than plain input (ANSI codes added); got len=%d want > %d", len(result), len(line))
+	}
+}
+
+// TestExtractText_EndOfLineSentinel mirrors TestHighlightLine_EndOfLineSentinel
+// for the extraction path used by the clipboard copy.
+func TestExtractText_EndOfLineSentinel(t *testing.T) {
+	line := "Hello, World!"
+	got := ExtractText(line, 7, -1)
+	want := "World!"
+	if got != want {
+		t.Errorf("ExtractText(line, 7, -1) = %q, want %q", got, want)
+	}
+}
+
+// TestIsLineInRange_StartLineSentinelHighlights composes IsLineInRange with
+// HighlightLine end-to-end: the start line of a multi-line, single-item
+// selection must actually emit highlight ANSI codes. This is the contract
+// the rendering path in scrolllist.View() relies on.
+func TestIsLineInRange_StartLineSentinelHighlights(t *testing.T) {
+	r := Range{StartItemIdx: 5, EndItemIdx: 5, StartLine: 0, EndLine: 2, StartCol: 0, EndCol: 10}
+	inRange, sc, ec := IsLineInRange(r, 5, 0)
+	if !inRange {
+		t.Fatalf("item 5 line 0 should be in range")
+	}
+	highlighted := HighlightLine("first line of selection", sc, ec)
+	if highlighted == "first line of selection" {
+		t.Errorf("first line of multi-line selection was not highlighted (sc=%d ec=%d)", sc, ec)
+	}
+}
+
 // TestMultiClickDetection verifies the click counting logic.
 func TestMultiClickDetection(t *testing.T) {
 	s := NewState()
@@ -243,7 +243,7 @@ host.ClearSession()

 ## Re-exported Types

-The SDK re-exports types so you don't need direct internal imports:
+The SDK re-exports message/session/MCP types so you don't need direct internal imports. Agent-configuration types are Kit-owned (not aliases) and use only SDK types in their signatures, so consumers never need to import the underlying LLM-provider package.

 ```go
 // Message types
@@ -251,13 +251,28 @@ kit.Message, kit.MessageRole, kit.ContentPart
 kit.TextContent, kit.ReasoningContent, kit.ToolCall, kit.ToolResult, kit.Finish
 kit.RoleUser, kit.RoleAssistant, kit.RoleTool, kit.RoleSystem

-// LLM types — concrete Kit-owned structs, no external library dependency
+// LLM types — Kit-owned `LLM*` aliases over the underlying provider types,
+// so consumers never import the provider package directly
 kit.LLMMessage      // {Role LLMMessageRole, Content string}
 kit.LLMMessageRole  // "user" | "assistant" | "system" | "tool"
 kit.LLMUsage        // {InputTokens, OutputTokens, TotalTokens, ...}
 kit.LLMResponse     // {Content, FinishReason, Usage}
 kit.LLMFilePart     // {Filename, Data []byte, MediaType}

+// Agent configuration — concrete Kit-owned structs and function types.
+// All fields use SDK types (e.g. `[]kit.Tool`), so consumers can construct
+// these without importing any LLM-provider package.
+kit.AgentConfig              // Lower-level agent config — prefer Options unless you need direct control
+kit.DebugLogger              // Interface: LogDebug(string) / IsDebugEnabled() bool
+kit.MCPTaskConfig            // Task-aware MCP tools/call config (modes, polling, progress)
+kit.ToolCallHandler          // func(toolCallID, toolName, toolArgs string)
+kit.ToolExecutionHandler     // func(toolCallID, toolName, toolArgs string, isStarting bool)
+kit.ToolResultHandler        // func(toolCallID, toolName, toolArgs, result, metadata string, isError bool)
+kit.ResponseHandler          // func(content string)
+kit.StreamingResponseHandler // func(content string)
+kit.ToolCallContentHandler   // func(content string)
+kit.SpinnerFunc              // func(fn func() error) error
+
 // MCP OAuth types
 kit.MCPServer            // *server.MCPServer for in-process MCP transport
 kit.MCPServerConfig      // Configuration for an MCP server (stdio, SSE, or in-process)
@@ -0,0 +1,208 @@
+package kit
+
+import (
+	"context"
+	"errors"
+	"testing"
+	"time"
+
+	"github.com/mark3labs/kit/internal/agent"
+)
+
+// TestAgentConfigToInternal verifies that the SDK-side AgentConfig converts
+// faithfully to the internal agent.AgentConfig representation, preserving
+// every field consumed by the internal agent layer.
+//
+// Regression test for https://github.com/mark3labs/kit/issues/30.
+func TestAgentConfigToInternal(t *testing.T) {
+	t.Run("nil receiver returns nil", func(t *testing.T) {
+		var c *AgentConfig
+		if got := c.toInternal(); got != nil {
+			t.Errorf("nil.toInternal() = %v, want nil", got)
+		}
+	})
+
+	t.Run("scalar fields round-trip", func(t *testing.T) {
+		c := &AgentConfig{
+			SystemPrompt:     "sys",
+			MaxSteps:         7,
+			StreamingEnabled: true,
+			DisableCoreTools: true,
+		}
+		got := c.toInternal()
+		if got == nil {
+			t.Fatal("toInternal() = nil")
+		}
+		if got.SystemPrompt != "sys" {
+			t.Errorf("SystemPrompt = %q, want %q", got.SystemPrompt, "sys")
+		}
+		if got.MaxSteps != 7 {
+			t.Errorf("MaxSteps = %d, want 7", got.MaxSteps)
+		}
+		if !got.StreamingEnabled {
+			t.Error("StreamingEnabled = false, want true")
+		}
+		if !got.DisableCoreTools {
+			t.Error("DisableCoreTools = false, want true")
+		}
+	})
+
+	t.Run("tool slices propagate without conversion", func(t *testing.T) {
+		// Tool is a type alias for the underlying LLM-tool type, so the
+		// SDK []Tool and internal []fantasy.AgentTool slices share the
+		// same backing array after conversion.
+		tool := NewTool[struct{}]("noop", "noop", nil)
+		c := &AgentConfig{
+			CoreTools:  []Tool{tool},
+			ExtraTools: []Tool{tool, tool},
+		}
+		got := c.toInternal()
+		if len(got.CoreTools) != 1 {
+			t.Errorf("CoreTools len = %d, want 1", len(got.CoreTools))
+		}
+		if len(got.ExtraTools) != 2 {
+			t.Errorf("ExtraTools len = %d, want 2", len(got.ExtraTools))
+		}
+	})
+
+	t.Run("tool wrapper is invoked through internal config", func(t *testing.T) {
+		called := false
+		c := &AgentConfig{
+			ToolWrapper: func(in []Tool) []Tool {
+				called = true
+				return in
+			},
+		}
+		got := c.toInternal()
+		if got.ToolWrapper == nil {
+			t.Fatal("internal ToolWrapper is nil")
+		}
+		_ = got.ToolWrapper(nil)
+		if !called {
+			t.Error("SDK ToolWrapper was not invoked through the internal config")
+		}
+	})
+
+	t.Run("OnMCPServerLoaded propagates", func(t *testing.T) {
+		var captured string
+		wantErr := errors.New("boom")
+		c := &AgentConfig{
+			OnMCPServerLoaded: func(name string, _ int, _ error) {
+				captured = name
+			},
+		}
+		got := c.toInternal()
+		got.OnMCPServerLoaded("svr", 3, wantErr)
+		if captured != "svr" {
+			t.Errorf("OnMCPServerLoaded captured = %q, want %q", captured, "svr")
+		}
+	})
+
+	t.Run("DebugLogger propagates", func(t *testing.T) {
+		dl := &fakeDebugLogger{enabled: true}
+		c := &AgentConfig{DebugLogger: dl}
+		got := c.toInternal()
+		if got.DebugLogger == nil {
+			t.Fatal("internal DebugLogger is nil")
+		}
+		if !got.DebugLogger.IsDebugEnabled() {
+			t.Error("IsDebugEnabled = false, want true")
+		}
+		got.DebugLogger.LogDebug("hello")
+		if len(dl.messages) != 1 || dl.messages[0] != "hello" {
+			t.Errorf("messages = %v, want [hello]", dl.messages)
+		}
+	})
+
+	t.Run("MCPTaskConfig propagates with mode + progress", func(t *testing.T) {
+		c := &AgentConfig{
+			MCPTaskConfig: MCPTaskConfig{
+				PerServerMode: map[string]MCPTaskMode{
+					"build-svr": MCPTaskModeAlways,
+				},
+				DefaultTTL:      30 * time.Second,
+				PollInterval:    250 * time.Millisecond,
+				MaxPollInterval: 2 * time.Second,
+				Timeout:         5 * time.Minute,
+				Progress:        func(_ MCPTaskProgress) {},
+			},
+		}
+		got := c.toInternal()
+		if got.MCPTaskConfig.DefaultTTL != 30*time.Second {
+			t.Errorf("DefaultTTL = %v, want 30s", got.MCPTaskConfig.DefaultTTL)
+		}
+		if got.MCPTaskConfig.PollInterval != 250*time.Millisecond {
+			t.Errorf("PollInterval = %v, want 250ms", got.MCPTaskConfig.PollInterval)
+		}
+		if got.MCPTaskConfig.MaxPollInterval != 2*time.Second {
+			t.Errorf("MaxPollInterval = %v, want 2s", got.MCPTaskConfig.MaxPollInterval)
+		}
+		if got.MCPTaskConfig.Timeout != 5*time.Minute {
+			t.Errorf("Timeout = %v, want 5m", got.MCPTaskConfig.Timeout)
+		}
+		mode, ok := got.MCPTaskConfig.PerServerMode["build-svr"]
+		if !ok {
+			t.Fatal("PerServerMode missing 'build-svr'")
+		}
+		if string(mode) != string(MCPTaskModeAlways) {
+			t.Errorf("mode = %q, want %q", mode, MCPTaskModeAlways)
+		}
+		if got.MCPTaskConfig.Progress == nil {
+			t.Fatal("internal Progress handler is nil")
+		}
+	})
+
+	t.Run("auth and token store factories are wired", func(t *testing.T) {
+		auth := &fakeAuthHandler{}
+		tokenCalls := 0
+		var tokenServer string
+		factory := MCPTokenStoreFactory(func(server string) (MCPTokenStore, error) {
+			tokenCalls++
+			tokenServer = server
+			return nil, nil
+		})
+		c := &AgentConfig{
+			AuthHandler:       auth,
+			TokenStoreFactory: factory,
+		}
+		got := c.toInternal()
+		if got.AuthHandler == nil {
+			t.Fatal("internal AuthHandler is nil")
+		}
+		if got.TokenStoreFactory == nil {
+			t.Fatal("internal TokenStoreFactory is nil")
+		}
+		_, _ = got.TokenStoreFactory("https://example.test")
+		if tokenCalls != 1 {
+			t.Errorf("token factory call count = %d, want 1", tokenCalls)
+		}
+		if tokenServer != "https://example.test" {
+			t.Errorf("token factory server arg = %q", tokenServer)
+		}
+		if got.AuthHandler.RedirectURI() != "redirect" {
+			t.Errorf("RedirectURI = %q, want %q", got.AuthHandler.RedirectURI(), "redirect")
+		}
+	})
+
+	// Compile-time check that the internal type is what we expect.
+	//nolint:staticcheck // QF1011: explicit type asserts the conversion target.
+	var _ *agent.AgentConfig = (&AgentConfig{}).toInternal()
+}
+
+// fakeAuthHandler implements both kit.MCPAuthHandler and the structurally
+// identical tools.MCPAuthHandler used by the internal layer.
+type fakeAuthHandler struct{}
+
+func (f *fakeAuthHandler) RedirectURI() string { return "redirect" }
+func (f *fakeAuthHandler) HandleAuth(_ context.Context, _ string, _ string) (string, error) {
+	return "", nil
+}
+
+// fakeDebugLogger implements kit.DebugLogger for tests.
+type fakeDebugLogger struct {
+	enabled  bool
+	messages []string
+}
+
+func (f *fakeDebugLogger) LogDebug(m string)    { f.messages = append(f.messages, m) }
+func (f *fakeDebugLogger) IsDebugEnabled() bool { return f.enabled }
@@ -76,6 +76,22 @@ type ExtensionAPI interface {
 	// Lifecycle
 	Reload() error
 	HasExtensions() bool
+
+	// Loaded returns metadata about the extensions currently loaded.
+	Loaded() []ExtensionInfo
+}
+
+// ExtensionInfo describes a single loaded extension for display purposes
+// (e.g. the startup banner or `kit extensions list`).
+type ExtensionInfo struct {
+	// Path is the absolute path of the extension's .go file.
+	Path string
+	// ToolCount is the number of tools registered by the extension.
+	ToolCount int
+	// CommandCount is the number of slash commands registered.
+	CommandCount int
+	// HandlerCount is the total number of event handlers registered.
+	HandlerCount int
 }

 // extensionAPI implements ExtensionAPI by wrapping a Kit instance.
@@ -456,3 +472,27 @@ func (e *extensionAPI) Reload() error {
 func (e *extensionAPI) HasExtensions() bool {
 	return e.kit.extRunner != nil
 }
+
+func (e *extensionAPI) Loaded() []ExtensionInfo {
+	if e.kit.extRunner == nil {
+		return nil
+	}
+	exts := e.kit.extRunner.Extensions()
+	if len(exts) == 0 {
+		return nil
+	}
+	infos := make([]ExtensionInfo, 0, len(exts))
+	for _, ex := range exts {
+		handlerCount := 0
+		for _, hs := range ex.Handlers {
+			handlerCount += len(hs)
+		}
+		infos = append(infos, ExtensionInfo{
+			Path:         ex.Path,
+			ToolCount:    len(ex.Tools),
+			CommandCount: len(ex.Commands),
+			HandlerCount: handlerCount,
+		})
+	}
+	return infos
+}
@@ -58,6 +58,9 @@ type Kit struct {
 	// When false, per-model system prompts from modelSettings/customModels
 	// can replace the default prompt on model switch.
 	hasCustomSystemPrompt bool
+	// systemPromptSource holds the raw configured value (file path or text)
+	// when hasCustomSystemPrompt is true; empty when the built-in default is in use.
+	systemPromptSource string

 	// Hook registries — interception layer (see hooks.go).
 	beforeToolCall  *hookRegistry[BeforeToolCallHook, BeforeToolCallResult]
@@ -632,6 +635,21 @@ func (m *Kit) SetModel(ctx context.Context, modelString string) error {
 	return nil
 }

+// HasCustomSystemPrompt reports whether the user explicitly configured a system
+// prompt via --system-prompt, a config file entry, or SDK Options.SystemPrompt.
+// When false, the built-in default (or a per-model override) is in use and can
+// be replaced transparently on model switch.
+func (m *Kit) HasCustomSystemPrompt() bool {
+	return m.hasCustomSystemPrompt
+}
+
+// GetSystemPromptSource returns the raw configured value — a file path or
+// inline text — when HasCustomSystemPrompt is true; returns an empty string
+// when the built-in default prompt is active.
+func (m *Kit) GetSystemPromptSource() string {
+	return m.systemPromptSource
+}
+
 // composeSystemPrompt takes a base system prompt and composes it with the
 // current runtime context: AGENTS.md content, skills metadata, and date/cwd.
 // This mirrors the composition done during Kit.New() initialization.
@@ -1179,6 +1197,7 @@ func New(ctx context.Context, opts *Options) (*Kit, error) {
 		maxSteps              int
 		streaming             bool
 		hasCustomSystemPrompt bool
+		systemPromptSource    string
 	)

 	if err := func() error {
@@ -1285,13 +1304,27 @@ func New(ctx context.Context, opts *Options) (*Kit, error) {
 		// explicitly set system-prompt, use the per-model prompt as the
 		// base instead of the global default.
 		{
-			basePrompt := viper.GetString("system-prompt")
+			rawPromptInput := viper.GetString("system-prompt")
+
+			// Resolve a file path to its content so PromptBuilder receives the
+			// actual prompt text rather than a literal path string. Without this,
+			// when system-prompt is set to a file path in the config file or via
+			// --system-prompt, the path itself becomes the effective system prompt
+			// sent to the model (LoadSystemPrompt only ran later, after viper had
+			// been overwritten with the augmented base text).
+			basePrompt, _ := config.LoadSystemPrompt(rawPromptInput)
+			if basePrompt == "" {
+				basePrompt = rawPromptInput
+			}

 			// Track whether the user explicitly configured a custom system
 			// prompt. When they haven't (basePrompt is the built-in default
 			// or empty), per-model system prompts can replace it on switch.
 			userSetSystemPrompt := basePrompt != "" && basePrompt != defaultSystemPrompt
 			hasCustomSystemPrompt = userSetSystemPrompt
+			if hasCustomSystemPrompt {
+				systemPromptSource = rawPromptInput
+			}

 			// Check for per-model system prompt override when no explicit
 			// global system-prompt was configured by the user.
@@ -1456,7 +1489,7 @@ func New(ctx context.Context, opts *Options) (*Kit, error) {

 	if opts.CLI != nil {
 		setupOpts.ShowSpinner = opts.CLI.ShowSpinner
-		setupOpts.SpinnerFunc = opts.CLI.SpinnerFunc
+		setupOpts.SpinnerFunc = agent.SpinnerFunc(opts.CLI.SpinnerFunc)
 		setupOpts.UseBufferedLogger = opts.CLI.UseBufferedLogger
 		if opts.CLI.ProgressReaderFunc != nil {
 			providerConfig.ProgressReaderFunc = opts.CLI.ProgressReaderFunc
@@ -1500,6 +1533,7 @@ func New(ctx context.Context, opts *Options) (*Kit, error) {
 		opts:                  opts,
 		mcpConfig:             mcpConfig,
 		hasCustomSystemPrompt: hasCustomSystemPrompt,
+		systemPromptSource:    systemPromptSource,
 		beforeToolCall:        beforeToolCall,
 		afterToolResult:       afterToolResult,
 		beforeTurn:            beforeTurn,
@@ -3,6 +3,7 @@ package kit_test
 import (
 	"context"
 	"os"
+	"strings"
 	"testing"

 	"github.com/spf13/viper"
@@ -306,3 +307,92 @@ func TestSessionManagement(t *testing.T) {
 // resetViper wipes viper's global state so a test case doesn't leak
 // viper.Set() calls into the next one. Used via defer in subtests.
 func resetViper() { viper.Reset() }
+
+// TestNewSystemPromptFilePath is a regression test for issue #25.
+//
+// When Options.SystemPrompt (or the --system-prompt flag / config entry) is a
+// file path, Kit must resolve the path to its file contents *before* the
+// PromptBuilder composes the runtime context. Previously the path string
+// itself was used verbatim as the base prompt, so the LLM received the path —
+// not the prompt — as its system message.
+func TestNewSystemPromptFilePath(t *testing.T) {
+	if os.Getenv("ANTHROPIC_API_KEY") == "" {
+		t.Skip("Skipping test: ANTHROPIC_API_KEY not set")
+	}
+	defer resetViper()
+
+	const promptContent = "You are a strict regression-test persona. Marker: KIT-25-OK"
+
+	tmpFile, err := os.CreateTemp(t.TempDir(), "kit-system-prompt-*.md")
+	if err != nil {
+		t.Fatalf("failed to create temp prompt file: %v", err)
+	}
+	if _, err := tmpFile.WriteString(promptContent); err != nil {
+		t.Fatalf("failed to write temp prompt file: %v", err)
+	}
+	if err := tmpFile.Close(); err != nil {
+		t.Fatalf("failed to close temp prompt file: %v", err)
+	}
+
+	ctx := context.Background()
+	host, err := kit.New(ctx, &kit.Options{
+		Model:        "anthropic/claude-sonnet-4-5-20250929",
+		SystemPrompt: tmpFile.Name(),
+		Quiet:        true,
+		NoSession:    true,
+	})
+	if err != nil {
+		t.Fatalf("Failed to create Kit with system-prompt file: %v", err)
+	}
+	defer func() { _ = host.Close() }()
+
+	if !host.HasCustomSystemPrompt() {
+		t.Error("HasCustomSystemPrompt() = false; want true when --system-prompt is set")
+	}
+	if got, want := host.GetSystemPromptSource(), tmpFile.Name(); got != want {
+		t.Errorf("GetSystemPromptSource() = %q; want %q", got, want)
+	}
+
+	// The composed system prompt is written back to viper after PromptBuilder
+	// runs. It must contain the file's contents, not the file path.
+	composed := viper.GetString("system-prompt")
+	if !strings.Contains(composed, promptContent) {
+		t.Errorf("composed system-prompt does not contain file contents\n  composed = %q\n  want substring = %q", composed, promptContent)
+	}
+	if strings.TrimSpace(composed) == tmpFile.Name() {
+		t.Errorf("composed system-prompt is the file path verbatim (%q); LoadSystemPrompt was not applied before PromptBuilder", composed)
+	}
+}
+
+// TestNewSystemPromptInline confirms that inline system-prompt strings still
+// flow through unchanged after the file-path resolution change.
+func TestNewSystemPromptInline(t *testing.T) {
+	if os.Getenv("ANTHROPIC_API_KEY") == "" {
+		t.Skip("Skipping test: ANTHROPIC_API_KEY not set")
+	}
+	defer resetViper()
+
+	const inline = "You are a concise inline-prompt persona."
+
+	ctx := context.Background()
+	host, err := kit.New(ctx, &kit.Options{
+		Model:        "anthropic/claude-sonnet-4-5-20250929",
+		SystemPrompt: inline,
+		Quiet:        true,
+		NoSession:    true,
+	})
+	if err != nil {
+		t.Fatalf("Failed to create Kit with inline system-prompt: %v", err)
+	}
+	defer func() { _ = host.Close() }()
+
+	if !host.HasCustomSystemPrompt() {
+		t.Error("HasCustomSystemPrompt() = false; want true for inline prompt")
+	}
+	if got := host.GetSystemPromptSource(); got != inline {
+		t.Errorf("GetSystemPromptSource() = %q; want %q", got, inline)
+	}
+	if composed := viper.GetString("system-prompt"); !strings.Contains(composed, inline) {
+		t.Errorf("composed system-prompt missing inline content; got %q", composed)
+	}
+}
@@ -98,6 +98,70 @@ type MCPTaskProgress struct {
 // dispatched on a goroutine.
 type MCPTaskProgressHandler func(MCPTaskProgress)

+// MCPTaskConfig configures task-aware MCP tools/call execution. All fields
+// are optional; the zero value disables progress callbacks and applies
+// sensible polling defaults inside the engine.
+//
+// For most consumers, the flat [Options] fields (`MCPTaskMode`,
+// `MCPTaskTTL`, `MCPTaskPollInterval`, `MCPTaskMaxPollInterval`,
+// `MCPTaskTimeout`, `MCPTaskProgress`) are the preferred entry point.
+// MCPTaskConfig is exposed for the low-level [AgentConfig] path.
+type MCPTaskConfig struct {
+	// PerServerMode overrides the per-server task mode resolved from
+	// [MCPServerConfig]. Keys are server names. Missing entries fall back
+	// to the configured value.
+	PerServerMode map[string]MCPTaskMode
+
+	// DefaultTTL is the TTL hint sent in TaskParams when augmenting a
+	// tools/call. Zero means omit the TTL — let the server pick its own.
+	DefaultTTL time.Duration
+
+	// PollInterval is the fallback interval between tasks/get requests
+	// when the server does not suggest one. Zero defaults to 1 second.
+	PollInterval time.Duration
+
+	// MaxPollInterval caps the polling interval. Zero defaults to 5 seconds.
+	MaxPollInterval time.Duration
+
+	// Timeout is the maximum wall-clock duration to wait for a task to
+	// reach a terminal state. Zero defaults to 15 minutes. Independent
+	// of the per-call context deadline; whichever fires first wins.
+	Timeout time.Duration
+
+	// Progress, if non-nil, receives every status transition observed by
+	// the polling loop.
+	Progress MCPTaskProgressHandler
+}
+
+// toToolsConfig converts the SDK-level [MCPTaskConfig] to the internal
+// tools-package representation. Keeps the dependency arrow internal-only.
+func (c MCPTaskConfig) toToolsConfig() tools.MCPTaskConfig {
+	cfg := tools.MCPTaskConfig{
+		DefaultTTL:      c.DefaultTTL,
+		PollInterval:    c.PollInterval,
+		MaxPollInterval: c.MaxPollInterval,
+		Timeout:         c.Timeout,
+	}
+	if len(c.PerServerMode) > 0 {
+		cfg.PerServerMode = make(map[string]tools.MCPTaskMode, len(c.PerServerMode))
+		for k, v := range c.PerServerMode {
+			cfg.PerServerMode[k] = tools.MCPTaskMode(v)
+		}
+	}
+	if c.Progress != nil {
+		h := c.Progress
+		cfg.Progress = func(p tools.MCPTaskProgress) {
+			h(MCPTaskProgress{
+				Server:  p.Server,
+				TaskID:  p.TaskID,
+				Status:  MCPTaskStatus(p.Status),
+				Message: p.Message,
+			})
+		}
+	}
+	return cfg
+}
+
 // mcpTaskOptions carries SDK consumer configuration into the agent setup.
 // Stored on Options as a single value so the public surface stays compact;
 // individual fields are exposed via WithMCP* builder functions.
@@ -11,6 +11,7 @@ import (
 	"github.com/mark3labs/kit/internal/message"
 	"github.com/mark3labs/kit/internal/models"
 	"github.com/mark3labs/kit/internal/session"
+	"github.com/mark3labs/kit/internal/tools"
 	"github.com/mark3labs/mcp-go/client/transport"
 	"github.com/mark3labs/mcp-go/server"
 )
@@ -75,25 +76,151 @@ type Config = config.Config
 // local (stdio) and remote (StreamableHTTP/SSE) server types.
 type MCPServerConfig = config.MCPServerConfig

-// ==== Agent Types (internal/agent/) ====
+// ==== Agent Types ====

-// AgentConfig holds configuration options for creating a new Agent.
-type AgentConfig = agent.AgentConfig
+// DebugLogger is an SDK-owned interface for low-level debug logging from
+// the engine and MCP tool plumbing. Implementations must be safe for
+// concurrent use.
+//
+// Most consumers do not need to provide one; pass [Options.Debug] = true
+// to use the default logger. DebugLogger is exposed for the low-level
+// [AgentConfig] path and for embedders that want to route debug output
+// into their own logging system.
+type DebugLogger interface {
+	// LogDebug records a single debug message. Implementations may drop,
+	// buffer, or render the message however they choose.
+	LogDebug(message string)
+	// IsDebugEnabled reports whether debug logging is active. Callers may
+	// check this before doing expensive formatting work.
+	IsDebugEnabled() bool
+}

-type (
-	// ToolCallHandler is a function type for handling tool calls as they happen.
-	ToolCallHandler = agent.ToolCallHandler
-	// ToolExecutionHandler is a function type for handling tool execution start/end events.
-	ToolExecutionHandler = agent.ToolExecutionHandler
-	// ToolResultHandler is a function type for handling tool results.
-	ToolResultHandler = agent.ToolResultHandler
-	// ResponseHandler is a function type for handling LLM responses.
-	ResponseHandler = agent.ResponseHandler
-	// StreamingResponseHandler is a function type for handling streaming LLM responses.
-	StreamingResponseHandler = agent.StreamingResponseHandler
-	// ToolCallContentHandler is a function type for handling content that accompanies tool calls.
-	ToolCallContentHandler = agent.ToolCallContentHandler
-)
+// AgentConfig holds configuration options for constructing an agent at the
+// SDK boundary. All fields use SDK-owned types, so consumers can populate
+// this struct without importing any underlying LLM-provider package.
+//
+// For most use cases, prefer the high-level [New] entry point with
+// [Options]. AgentConfig is exposed for advanced consumers that need
+// direct access to the lower-level agent configuration shape.
+type AgentConfig struct {
+	// ModelConfig holds the LLM provider configuration. A nil value means
+	// that the default provider/model resolution will be used.
+	ModelConfig *ProviderConfig
+
+	// MCPConfig describes any MCP servers whose tools should be loaded
+	// alongside core tools.
+	MCPConfig *Config
+
+	// SystemPrompt is the system prompt sent to the LLM.
+	SystemPrompt string
+
+	// MaxSteps caps the number of LLM iterations per turn. A value of
+	// zero means no cap is applied at this layer.
+	MaxSteps int
+
+	// StreamingEnabled controls whether the agent streams responses.
+	StreamingEnabled bool
+
+	// AuthHandler handles OAuth authorization for remote MCP servers.
+	// When nil, remote MCP servers requiring OAuth will fail to connect.
+	AuthHandler MCPAuthHandler
+
+	// TokenStoreFactory, if non-nil, creates a custom token store for each
+	// remote MCP server's OAuth tokens. When nil, the default file-based
+	// token store is used.
+	TokenStoreFactory MCPTokenStoreFactory
+
+	// CoreTools overrides the default core tool set. If empty, [AllTools]
+	// is used. Provide a custom tool set (e.g. [CodingTools] or tools
+	// built with a custom WorkDir) to scope agent capabilities.
+	CoreTools []Tool
+
+	// DisableCoreTools, when true, prevents loading any core tools.
+	// Combined with empty CoreTools this yields a chat-only agent with
+	// no built-in tools.
+	DisableCoreTools bool
+
+	// ExtraTools are additional tools loaded alongside core and MCP tools.
+	ExtraTools []Tool
+
+	// ToolWrapper, if non-nil, wraps the combined tool list before it is
+	// handed to the LLM. Used to intercept tool calls or results.
+	ToolWrapper func([]Tool) []Tool
+
+	// OnMCPServerLoaded, if non-nil, is invoked once for each MCP server
+	// when its tools have finished loading (or failed). Called from a
+	// background goroutine.
+	OnMCPServerLoaded func(serverName string, toolCount int, err error)
+
+	// DebugLogger receives low-level debug output from the engine and the
+	// MCP tool plumbing. Nil means no debug output is emitted at this
+	// layer (regardless of [Options.Debug], which feeds the higher-level
+	// [New] entry point). Pass an implementation here when wiring a custom
+	// logger through the lower-level AgentConfig path.
+	DebugLogger DebugLogger
+
+	// MCPTaskConfig configures task-aware MCP tools/call execution — mode
+	// overrides, polling intervals, timeouts, and the progress handler.
+	// The zero value preserves historical synchronous-only behaviour for
+	// any server that didn't advertise task support during initialize.
+	MCPTaskConfig MCPTaskConfig
+}
+
+// toInternal converts an AgentConfig to its internal representation.
+// Slice and function fields convert without allocation because [Tool]
+// is a type alias for the underlying LLM-tool type.
+func (c *AgentConfig) toInternal() *agent.AgentConfig {
+	if c == nil {
+		return nil
+	}
+	out := &agent.AgentConfig{
+		ModelConfig:       c.ModelConfig,
+		MCPConfig:         c.MCPConfig,
+		SystemPrompt:      c.SystemPrompt,
+		MaxSteps:          c.MaxSteps,
+		StreamingEnabled:  c.StreamingEnabled,
+		CoreTools:         c.CoreTools,
+		DisableCoreTools:  c.DisableCoreTools,
+		ExtraTools:        c.ExtraTools,
+		ToolWrapper:       c.ToolWrapper,
+		OnMCPServerLoaded: c.OnMCPServerLoaded,
+	}
+	if c.AuthHandler != nil {
+		out.AuthHandler = c.AuthHandler
+	}
+	if c.TokenStoreFactory != nil {
+		out.TokenStoreFactory = tools.TokenStoreFactory(c.TokenStoreFactory)
+	}
+	if c.DebugLogger != nil {
+		out.DebugLogger = c.DebugLogger
+	}
+	out.MCPTaskConfig = c.MCPTaskConfig.toToolsConfig()
+	return out
+}
+
+// ToolCallHandler is invoked when the LLM produces a tool call. It receives
+// the call ID, tool name, and the JSON-encoded input arguments.
+type ToolCallHandler func(toolCallID, toolName, toolArgs string)
+
+// ToolExecutionHandler is invoked at the start and end of tool execution.
+// The isStarting flag distinguishes the two phases.
+type ToolExecutionHandler func(toolCallID, toolName, toolArgs string, isStarting bool)
+
+// ToolResultHandler is invoked after a tool finishes executing. The metadata
+// parameter carries optional structured data (e.g. file-diff info) from the
+// tool execution, JSON-encoded; it may be empty.
+type ToolResultHandler func(toolCallID, toolName, toolArgs, result, metadata string, isError bool)
+
+// ResponseHandler is invoked with the final assistant text for each turn.
+type ResponseHandler func(content string)
+
+// StreamingResponseHandler is invoked with each streamed text delta as it
+// arrives from the LLM.
+type StreamingResponseHandler func(content string)
+
+// ToolCallContentHandler is invoked with any assistant text that accompanies
+// a tool call within the same step.
+type ToolCallContentHandler func(content string)

 // ==== Provider & Model Types (internal/models/) ====

@@ -126,7 +253,7 @@ type ModelsRegistry = models.ModelsRegistry

 // SpinnerFunc wraps a function in a loading spinner animation. Used for
 // Ollama model loading. Signature: func(fn func() error) error.
-type SpinnerFunc = agent.SpinnerFunc
+type SpinnerFunc func(fn func() error) error

 // ==== LLM Types ====
 //
@@ -1,6 +1,7 @@
 package kit_test

 import (
+	"context"
 	"encoding/json"
 	"testing"

@@ -263,6 +264,101 @@ func TestConvertFromLLMMessage(t *testing.T) {
 	}
 }

+// TestAgentConfigNoFantasyImport verifies AgentConfig can be populated with
+// every field — including CoreTools, ExtraTools, and ToolWrapper — using
+// only SDK-owned types. This test deliberately does not import
+// "charm.land/fantasy"; the package compiling at all is the proof that the
+// SDK no longer leaks the dependency name through AgentConfig.
+//
+// Regression test for https://github.com/mark3labs/kit/issues/30.
+func TestAgentConfigNoFantasyImport(t *testing.T) {
+	myTool := kit.NewTool[struct{}]("noop", "does nothing", func(_ context.Context, _ struct{}) (kit.ToolOutput, error) {
+		return kit.TextResult("ok"), nil
+	})
+
+	wrapperCalled := false
+	cfg := kit.AgentConfig{
+		SystemPrompt:     "you are a tester",
+		MaxSteps:         5,
+		StreamingEnabled: true,
+		CoreTools:        []kit.Tool{myTool},
+		ExtraTools:       []kit.Tool{myTool},
+		DisableCoreTools: false,
+		ToolWrapper: func(in []kit.Tool) []kit.Tool {
+			wrapperCalled = true
+			return in
+		},
+		OnMCPServerLoaded: func(_ string, _ int, _ error) {},
+	}
+
+	if cfg.SystemPrompt != "you are a tester" {
+		t.Errorf("SystemPrompt = %q, want %q", cfg.SystemPrompt, "you are a tester")
+	}
+	if cfg.MaxSteps != 5 {
+		t.Errorf("MaxSteps = %d, want 5", cfg.MaxSteps)
+	}
+	if !cfg.StreamingEnabled {
+		t.Error("StreamingEnabled = false, want true")
+	}
+	if len(cfg.CoreTools) != 1 {
+		t.Errorf("CoreTools len = %d, want 1", len(cfg.CoreTools))
+	}
+	if len(cfg.ExtraTools) != 1 {
+		t.Errorf("ExtraTools len = %d, want 1", len(cfg.ExtraTools))
+	}
+
+	// Exercise the wrapper to confirm the func type is usable.
+	out := cfg.ToolWrapper(cfg.CoreTools)
+	if !wrapperCalled {
+		t.Error("ToolWrapper was not invoked")
+	}
+	if len(out) != 1 {
+		t.Errorf("wrapped tool list len = %d, want 1", len(out))
+	}
+}
+
+// TestAgentConfigToolWrapperSignature documents that AgentConfig.ToolWrapper
+// uses kit.Tool (not the underlying provider type) in its signature.
+func TestAgentConfigToolWrapperSignature(t *testing.T) {
+	//nolint:staticcheck // QF1011: explicit type asserts the SDK-side func signature.
+	var _ func([]kit.Tool) []kit.Tool = func(in []kit.Tool) []kit.Tool { return in }
+	cfg := kit.AgentConfig{
+		ToolWrapper: func(in []kit.Tool) []kit.Tool { return in },
+	}
+	if cfg.ToolWrapper == nil {
+		t.Fatal("ToolWrapper assignment failed")
+	}
+}
+
+// TestSpinnerFuncSignature verifies SpinnerFunc has the documented signature
+// and can be constructed without importing any provider package.
+func TestSpinnerFuncSignature(t *testing.T) {
+	called := false
+	var sp kit.SpinnerFunc = func(fn func() error) error {
+		called = true
+		return fn()
+	}
+	err := sp(func() error { return nil })
+	if err != nil {
+		t.Errorf("SpinnerFunc returned err: %v", err)
+	}
+	if !called {
+		t.Error("SpinnerFunc did not invoke fn")
+	}
+}
+
+// TestHandlerTypesSignatures verifies the SDK-owned handler function types
+// can be assigned from plain function literals using only standard library
+// types in their signatures (no provider-package import required).
+func TestHandlerTypesSignatures(t *testing.T) {
+	var _ kit.ToolCallHandler = func(_, _, _ string) {}
+	var _ kit.ToolExecutionHandler = func(_, _, _ string, _ bool) {}
+	var _ kit.ToolResultHandler = func(_, _, _, _, _ string, _ bool) {}
+	var _ kit.ResponseHandler = func(_ string) {}
+	var _ kit.StreamingResponseHandler = func(_ string) {}
+	var _ kit.ToolCallContentHandler = func(_ string) {}
+}
+
 // containsStr is a tiny helper to avoid importing strings in test.
 func containsStr(s, substr string) bool {
 	return len(s) >= len(substr) && (s == substr || len(s) > 0 && indexStr(s, substr) >= 0)
@@ -1,5 +0,0 @@
-# Specs
-
-| Spec | Status | Description |
-|------|--------|-------------|
-| [unified-bubbletea-architecture](unified-bubbletea-architecture.md) | Draft | Replace micro-program pattern with single Bubble Tea program + thick app layer |
Author	SHA1	Message	Date
Ed Zynda	488b0ad69f	fix(agent): track tool call args per ToolCallID for parallel calls (#33 ) Previously GenerateWithCallbacks stored the most recent tool call's args in a single shared variable, which got clobbered when a provider emitted multiple tool_use blocks in a single step. Every OnToolResult callback then received the args of the last OnToolCall, regardless of which call it was actually resolving — breaking any downstream UI, log, or trace that derived its description from the toolArgs parameter. - Replace the shared currentToolArgs with a map keyed by ToolCallID, guarded by a sync.Mutex in case the streaming layer dispatches callbacks from multiple goroutines. - Delete each entry in OnToolResult so the map cannot accumulate across steps. - Add a regression test driving the streaming wrapper with a fake fantasy.Agent that emits two parallel tool calls before either result, asserting each callback sees its own args. Fixes #33	2026-05-20 10:23:29 +03:00
Ed Zynda	592f8dc84f	chore(models): refresh embedded models.dev snapshot - Sync internal/models/embedded_models.json from https://models.dev/api.json - Providers: 114 → 131 (+17) - Models: 4276 → 4817 (+541)	2026-05-19 15:11:01 +03:00
Ed Zynda	66c4a1eb15	build(deps): bump all dependencies and go directive to 1.26.3 - charm.land/fantasy v0.23.0 -> v0.25.0 - charmbracelet/ultraviolet -> 20260511 - coder/acp-go-sdk v0.12.2 -> v0.13.0 - mark3labs/mcp-go v0.51.0 -> v0.54.0 - kaptinlin/{go-i18n,jsonpointer,jsonschema,messageformat-go} bumps - golang.org/x/{crypto,net,sys,term,text} minor bumps - google.golang.org/{api,genai,genproto,grpc} bumps - charmbracelet/x/exp/{charmtone,slice}, tidwall/gjson, others - go directive bumped to 1.26.3 (required by fantasy v0.25.0) No code changes required; build, vet, and race tests all pass.	2026-05-19 13:24:53 +03:00
Ed Zynda	5104477631	perf(session): parallelize session list extraction Open the /resume session picker faster by extracting per-file metadata across a GOMAXPROCS-sized worker pool instead of sequentially. Each extractSessionInfo call is I/O + JSON-parse bound and independent, so wall time drops roughly proportionally to core count — meaningful for users with many sessions, where ListSessions + ListAllSessions ran back-to-back on the UI goroutine before the picker rendered.	2026-05-16 16:19:38 +03:00
Ed Zynda	394a4676a1	fix(app): deliver trailing widget update so layout resets after removal - Switch NotifyWidgetUpdate from leading-only to leading+trailing edge coalescing so a rapid SetWidget→RemoveWidget pair (e.g. emitted by subagent-monitor on SubagentEnd) is never silently dropped. - Without the trailing send the TUI keeps the pre-removal widget height, leaving empty rows below the status bar until some other event re-renders the layout.	2026-05-16 14:07:58 +03:00
Ed Zynda	30f2bc243d	fix(ui): correct mouse selection drift with extension widgets - Match View() and getItemAndLineAtY() row counts for empty items so streaming-reasoning placeholders no longer offset hit-testing by one row each (exposed when extension widgets like subagent-monitor shrink the scrollback). - Honor IsLineInRange's endCol=-1 'to end of line' sentinel in HighlightLine and ExtractText so the start row of a multi-line drag actually renders highlighted and is included in clipboard copies. - Add regression tests for both invariants in scrolllist and selection.	2026-05-16 13:48:51 +03:00
Ed Zynda	922e246098	feat(prompts): auto-reload prompts and extensions from XDG config - Add prompts.GlobalDir() resolving $XDG_CONFIG_HOME/kit/prompts/ (default ~/.config/kit/prompts/) so prompt templates live alongside extensions and skills under the same XDG-aligned root. - LoadAll now discovers templates from both the legacy ~/.kit/prompts/ and the XDG location; existing legacy paths keep precedence. - Include GlobalDir() in the prompts/skills file watcher so edits under ~/.config/kit/prompts/ hot-reload automatically. - Surface a visible 'Extensions reloaded.' (or error) message when the extension watcher fires, matching /reload-ext feedback. - Restore examples/extensions/subagent-monitor.go alongside its test and update the test load path; previous move left the test broken.	2026-05-15 14:31:51 +03:00
Ed Zynda	32b6376515	chore: move go-edit-lint extension to global scope - Remove .kit/extensions/go-edit-lint.go from the repo since the extension is now installed under ~/.config/kit/extensions/ for per-user use across all projects.	2026-05-15 14:18:57 +03:00
Ed Zynda	cf194ff89a	feat(ui): list loaded extensions in startup banner - Add ExtensionInfo type and Loaded() method to the public ExtensionAPI so SDK consumers can inspect which extensions are active. - Introduce ui.ExtensionItem and thread ExtensionItems/GetExtensionItems through AppModelOptions, mirroring the existing SkillItem pattern. - Render an [Extensions] row in AddStartupMessageToScrollList showing the filename of each loaded extension (with a (N tools) suffix when extensions register tools). Falls back to tool count only when items are unavailable, and is omitted entirely when no extensions load. - Refresh the list on /reload-ext via a new refreshExtensionItems hook so the banner stays accurate across hot-reloads. - Add buildExtensionItems helper in cmd/root.go that strips .go and resolves subdirectory extensions to their parent dir name, tagging each as project or user scope based on cwd.	2026-05-15 14:08:42 +03:00
Ed Zynda	03006425fa	cleanup	2026-05-15 13:55:32 +03:00
Ed Zynda	a322dfc59a	fix(ui): eliminate mouse copy-selection drift during streaming - Lock viewport scroll while a drag-select is active so highlighted content stays under the cursor (SetItems, appendStreamingChunk, MouseWheelDown all now honor IsMouseDown). - HandleMouseDrag defensively clears autoScroll on every update so a racy re-enable can't shift the row mid-drag. - Recompute scrollback yOffset/viewport height on each mouse event via currentScrollbackBounds() instead of relying on stale values cached during the previous View() pass. - Account for canceling/ctrlCPressedOnce warning rows in distributeHeight and mark layoutDirty when those flags toggle so the height budget and mouse origin stay in sync. - Add ScrollList regression tests covering the three invariants.	2026-05-15 13:30:57 +03:00
Ed Zynda	b1387d837e	feat(ui): add /copy slash command to copy last message - Register /copy (alias /cp) in the System command category - Walk the scrollback to find the last user/assistant/reasoning message, skipping transient system messages - Reuse internal/ui/clipboard.CopyToClipboard for OSC 52 + native clipboard support (works over SSH) - Document the command in /help	2026-05-15 13:06:35 +03:00
Ed Zynda	f561f4cfd9	fix(session): order kept messages before post-compact branch in BuildContext After /compact, BuildContext emitted [summary, post-compact, kept] which placed an older kept user/assistant turn after the latest post-compaction turn. This broke user/assistant alternation and caused the model to respond as if the post-compaction turn never happened on the next user message. - Emit kept messages chronologically before post-compaction messages - Mirror the same order in GetContextEntryIDs so cut-point to entry-ID mapping stays aligned across repeat compactions - Update TestCompactionWithNewMessagesAfterCompaction to assert the correct chronological order	2026-05-14 20:42:20 +03:00
Ed Zynda	64caed57d4	fix(sdk): stop leaking fantasy types through pkg/kit.AgentConfig (#30 ) (#32 ) * fix(sdk): stop leaking fantasy types through pkg/kit.AgentConfig (#30) Replace the alias-based AgentConfig and handler types with SDK-owned structs and function types. CoreTools / ExtraTools / ToolWrapper now accept []kit.Tool, and the handler types (ToolCallHandler, ToolExecutionHandler, ToolResultHandler, ResponseHandler, StreamingResponseHandler, ToolCallContentHandler) plus SpinnerFunc are declared in pkg/kit/ with signatures that reference only SDK types. Consumers no longer need to import charm.land/fantasy to populate an AgentConfig or assign a handler. go doc pkg/kit AgentConfig output no longer mentions fantasy.. - Add unexported (AgentConfig).toInternal() to convert at the SDK boundary; Tool is still an alias for the underlying tool type, so slice and function fields convert without allocation. - Add agent_config_internal_test.go covering nil receiver, scalar fields, tool slices, ToolWrapper invocation, OnMCPServerLoaded, and auth/token-factory wiring. - Add types_test.go cases that populate AgentConfig and SpinnerFunc without importing fantasy -- the file compiling is the regression proof for the leak. - Update pkg/kit/README.md Re-exported Types section to record that AgentConfig and the handler types are now Kit-owned. Fixes #30 * fix(sdk): add DebugLogger and MCPTaskConfig to kit.AgentConfig (#30) The first revision of the SDK-owned AgentConfig dropped two fields that internal/agent.AgentConfig carried: DebugLogger (tools.DebugLogger) and MCPTaskConfig (tools.MCPTaskConfig). Restore them with SDK-owned equivalents and wire them through toInternal(). - Add kit.DebugLogger interface (LogDebug / IsDebugEnabled) mirroring tools.DebugLogger. Interface-to-interface assignment is automatic because the method sets match. - Add kit.MCPTaskConfig struct mirroring tools.MCPTaskConfig with SDK types (MCPTaskMode, MCPTaskProgressHandler) and a toToolsConfig() helper that converts at the SDK boundary. - Wire both new fields in (*AgentConfig).toInternal(). - Extend agent_config_internal_test.go with cases for both fields. - Document the additions in pkg/kit/README.md.	2026-05-13 21:10:28 +03:00
Ed Zynda	975c30a773	fix(mcp): surface MCP tool failures as soft errors, not critical aborts (#31 ) The MCP adapter previously wrapped any error returned by MCPToolManager.ExecuteTool into a Go error returned from the fantasy.AgentTool.Run interface. The fantasy agent loop treats those as critical errors and aborts the entire turn — discarding all prior reasoning, tool calls, and results. In practice that meant a single misbehaved MCP server returning a JSON-RPC "-32602 Invalid params" (e.g. a Zod schema mismatch on the server's input validation) would kill an in-progress turn after the model had already done dozens of seconds of useful work, with no way for the model to see the validation message and self-correct. This mismatched the contract that native Kit tools follow: native tools return errors via kit.ErrorResult(...), which become soft tool-result errors that the model reads and can act on (retry with corrected args, try a different tool, give up gracefully). Make the MCP path behave the same way: - JSON-RPC protocol errors, transport failures, and server-side schema rejections are now returned as fantasy.NewTextErrorResponse(...) with err == nil, so the agent loop continues and the model sees the failure in-band as a tool result it can reason about. - Context cancellation (ctx.Err() != nil) remains a critical error so callers can abort turns deterministically. This is the only case where bubbling up is correct — the caller intentionally tore the turn down and the agent must not keep spinning. - Server-side soft errors (CallToolResult{ isError: true }) and the happy path are unchanged. The agent loop's MaxSteps cap already bounds the worst case for a permanently broken MCP server, so there is no risk of unbounded retries. Side effect: extracted a tiny mcpExecutor interface for the one method the adapter uses (ExecuteTool), purely so the adapter is unit-testable in isolation without standing up a full MCPToolManager + connection pool. Behavior change note for downstream consumers: code that relied on host.PromptResult / Stream returning a Go error containing "mcp tool execution failed" will no longer see those errors — the failure information is now in the assistant's final response (or in the OnAfterToolResult / OnToolResult hooks, where IsError will be true). Context cancellation continues to surface as an error from those calls as before. Co-authored-by: space_cowboy <space_cowboy@mark3labs.com>	2026-05-13 20:12:31 +03:00
Ed Zynda	35b9360d64	feat(ui): autocomplete /skill:<name> slash commands - register loaded skills into the input autocomplete under category "Skills" with HasArgs so Enter populates "/skill:name " instead of auto-submitting, leaving room for trailing args - prefix descriptions with [project] or [user] to disambiguate colliding skill names across sources - extend refreshSkillItems to prune & re-add Skills entries on ContentReloadEvent, matching the pattern used for prompt templates and MCP prompts - add Description field to ui.SkillItem and populate it from kit.Skill.Description in both initial build and hot-reload paths	2026-05-13 15:35:07 +03:00
Ed Zynda	1b8373e133	cleanup	2026-05-12 13:30:30 +03:00
Ed Zynda	1a5e4ce7c5	Merge pull request #29 from mark3labs/fix/27-queued-messages-after-compact test(app): cover steer-drain branch of releaseBusyAfterCompact	2026-05-08 13:11:45 +03:00
Ed Zynda	8823977612	test(app): cover steer-drain branch of releaseBusyAfterCompact - Add unexported steerDrainFn test seam on App so unit tests can inject fake steer items without standing up a full *kit.Kit (Options.Kit is a concrete struct, not an interface). - releaseBusyAfterCompact now prefers the seam over Kit.DrainSteer via a small switch; production behaviour is unchanged when the field is nil. - Add TestReleaseBusyAfterCompact_splicesSteerAheadOfQueue, which pre-populates both fake steer items and ordinary queue prompts, invokes releaseBusyAfterCompact, and asserts the first dispatched prompt is the steer item — proving steer messages retain 'act now' priority and that drainQueue is actually launched (the bug from #27).	2026-05-08 12:18:52 +03:00
Ed Zynda	24e2ea111c	Merge pull request #28 from mark3labs/fix/27-queued-messages-after-compact fix(app): flush queued messages after /compact completes (#27)	2026-05-08 12:16:28 +03:00
Ed Zynda	31ea80ec4f	fix(app): flush queued messages after /compact completes (#27 ) - Add releaseBusyAfterCompact() shared deferred tail used by both CompactConversation and CompactAsync. It drains the SDK steer channel, splices steer items in front of any queued prompts, and hands off to drainQueue so messages received during compaction are dispatched automatically once compaction finishes. - Previously, busy was simply cleared on completion and the queue sat idle until the user submitted another prompt, which then flushed everything together. - Honor the closed flag so a teardown during compaction discards pending items instead of spawning drainQueue against a torn-down App. - Add regression tests covering the queued-flush, idle-empty, and closed-during-compact paths. Fixes #27	2026-05-08 11:30:26 +03:00
Ed Zynda	99f2680c2e	Merge pull request #26 from mark3labs/fix/25-system-prompt-file-path fix(kit): resolve system-prompt file path before PromptBuilder (#25)	2026-05-08 10:54:09 +03:00
Ed Zynda	da7e05eb87	fix(cmd): nil-guard CLI when emitting system-prompt notice in quiet mode SetupCLIForNonInteractive returns nil when --quiet is active, matching the pre-existing nil checks elsewhere in the same block (e.g. the buffered debug-message branch). Without this guard the new 'System Prompt loaded' notice panicked on quiet, non-interactive runs. Discovered via tmux smoke test of the #25 fix.	2026-05-08 10:44:01 +03:00
Ed Zynda	a95714a22d	fix(kit): resolve system-prompt file path before PromptBuilder (#25 ) When system-prompt was a file path (via --system-prompt, config entry, or SDK Options.SystemPrompt), the path string itself was used as the base prompt because config.LoadSystemPrompt only ran later in BuildProviderConfig — by which point viper had been overwritten with the path-augmented composed text. The LLM received the path instead of the prompt contents. - Call config.LoadSystemPrompt on the raw viper value in New() before PromptBuilder composes runtime context (AGENTS.md / skills / date). - Add HasCustomSystemPrompt() and GetSystemPromptSource() so SDK callers can inspect prompt state without reaching into viper. - Display 'System Prompt loaded: <source>' at startup in CLI and TUI modes, paralleling the per-server 'MCP server loaded' notice. - Add regression tests covering both file-path and inline prompt paths. Fixes #25	2026-05-08 10:39:14 +03:00
Ed Zynda	c4a2b0f1a3	Merge pull request #24 from mark3labs/audit-cleanup refactor: remove dead code and consolidate duplicated extension wiring	2026-05-07 17:46:49 +03:00