fix(tui): review follow-up — /retry, /plan, ANSI truncation, caching

- /retry: use session['history'] instead of non-existent
  agent.conversation_history; truncate history at last user message
  to match CLI retry_last() behavior; add history_lock safety
- /plan: pass user instruction (arg) to build_plan_path instead of
  session_key; add runtime_note so agent knows where to save the plan
- ANSI tool results: render full text via <Ansi wrap=truncate-end>
  instead of slicing raw ANSI through compactPreview (which cuts
  mid-escape-sequence producing garbled output)
- Move _PENDING_INPUT_COMMANDS frozenset to module level
- Use get_skill_commands() (cached) instead of scan_skill_commands()
  (rescans disk) in slash.exec skill interception
- Add 3 retry tests: happy path with history truncation verification,
  empty history error, multipart content extraction
- Update test mock target from scan_skill_commands to get_skill_commands

ui-tui/src/components/messageLine.tsx

@@ -28,13 +28,14 @@ export const MessageLine = memo(function MessageLine({
   }
 
   if (msg.role === 'tool') {
-    const preview = compactPreview(hasAnsi(msg.text) ? stripAnsi(msg.text) : msg.text, Math.max(24, cols - 14)) ||
-      '(empty tool result)'
+    const maxChars = Math.max(24, cols - 14)
+    const stripped = hasAnsi(msg.text) ? stripAnsi(msg.text) : msg.text
+    const preview = compactPreview(stripped, maxChars) || '(empty tool result)'
 
     return (
       <Box alignSelf="flex-start" borderColor={t.color.dim} borderStyle="round" marginLeft={3} paddingX={1}>
         {hasAnsi(msg.text) ? (
-          <Ansi>{compactPreview(msg.text, Math.max(24, cols - 14)) || '(empty tool result)'}</Ansi>
+          <Text wrap="truncate-end"><Ansi>{msg.text}</Ansi></Text>
         ) : (
           <Text color={t.color.dim} wrap="truncate-end">
             {preview}