[Bug] LLM chat output center-justified with text truncation in v2.0.73

Bug Description
Version 2.0.73

Installation immediately began a new issue with all chat output from the LLM. Information text such as tool call text are unaffected, but most chat output appears to be trying to center justify suddenly and this is also causing a lot of missing text. Below is an example turn with the AI:

● Write(test-output-demo.txt)
  ⎿  Wrote 1 lines to test-output-demo.txt
     CLAUDE CODE OUTPUT TEST FILE
     Generated: December 19, 2025
     Version Being Tested: 2.0.73
     Purpose: Demonstrate output rendering issues
     ═══════════════════════════════════════════════════════════
     SECTION 1: ASCII ART AND SPECIAL CHARACTERS
     ═══════════════════════════════════════════════════════════
     Box drawing characters:





                                         nstrate the variety of outputs Claude Code generates:





                                                           urned 100+ files)
                                                         son with line numbers
                                                                results)

**Environment Info**
- Platform: win32
- Terminal: cursor
- Version: 2.0.73
- Feedback ID: 4e1c1f02-6055-44a6-b465-83632bc81643

**Errors**
```json
[{"error":"Error: Request was aborted.\n    at gT.makeRequest (file:///C:/Users/Drickon/AppData/Local/nvm/v20.19.5/node_modules/@anthropic-ai/claude-code/cli.js:846:3940)\n    at process.processTicksAndRejections (node:internal/process/task_queues:95:5)","timestamp":"2025-12-19T06:25:16.242Z"}]