[BUG] Impenetrable error message

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

what

(venv) ubuntu@ip-x-y-z-w:~/src/my-project$ claude
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
 Settings Error

 /home/ubuntu/src/my-project/.claude/settings.local.json
  └ permissions
    └ allow
      └ "Bash(~/src/some-machine-learning/venv/bin/python -c "
  import matplotlib
        └ pyplot as plt
  import numpy as np

  # Data from the latest staged gradient descent with global refinement
  target =
          └ 0
            └ 04111543, 0
              └ 01312898, 0
                └ 0320538, 0
                  └ 06337341, 0
                    └ 08276107, 0
                      └ 1244354, 0
                        └ 15889854, 0
                          └ 14149855, 0
                            └ 1491152, 0
                              └ 19361962
                                └
  realized =
                                  └ 0
                                    └ 03325738, 0
                                      └ 00619568, 0
                                        └ 053082, 0
                                          └ 08482047, 0
                                            └ 12587228, 0
                                              └ 20932075, 0
                                                └ 14463585, 0
                                                  └ 05690389, 0
                                                    └ 13294094, 0
                                                      └ 15297078
                                                        └

  segments = range(10)
  width = 0
                                                          └ 35
  x = np
                                                            └ arange(len(segments))

  plt
                                                              └ figure(figsize=(12, 8))
  plt
                                                                └ bar(x - width/2, target, width, label=''Target'', color=''steelblue'', alpha=0
                                                                  └ 8)
  plt
                                                                    └ bar(x + width/2, realized, width, label=''Realized'', color=''orange'', alpha=0
                                                                      └ 8)

  # Add percentage error annotations
  for i in range(len(segments)):
      error_pct = (realized
                                                                        └ i
                                                                          └  - target
                                                                            └ i
                                                                              └ ) / target
                                                                                └ i
                                                                                  └  * 100
      plt
                                                                                    └ annotate(f''{error_pct:+
                                                                                      └ 1f}%'',
                  xy=(i, max(target
                                                                                        └ i
                                                                                          └ , realized
                                                                                            └ i
                                                                                              └ ) + 0
                                                                                                └ 01),
                  ha=''center'', va=''bottom'', fontsize=9,
                  color=''red'' if abs(error_pct) > 20 else ''green'')

  plt
                                                                                                  └ xlabel(''Segment'')
  plt
                                                                                                    └ ylabel(''Proportion'')
  plt
                                                                                                      └ title(''Target vs Realized Distributions - Staged GD with Global Refinement\nFinal
  Loss: 0
                                                                                                        └ 019350 (50% improvement from previous run)'')
  plt
                                                                                                          └ legend()
  plt
                                                                                                            └ xticks(x, segments)
  plt
                                                                                                              └ grid(True, alpha=0
                                                                                                                └ 3)
  plt
                                                                                                                  └ tight_layout()
  plt
                                                                                                                    └ savefig(''/tmp/distributions_comparison_latest
                                                                                                                      └ png'', dpi=150, bbox_inches=''tight'')
  print(''Created /tmp/distributions_comparison_latest
                                                                                                                        └ png'')
  ")": Empty parentheses. Either specify a pattern or use just "Bash" without parentheses. Examples: Bash, Bash(some-pattern)

 Files with errors are skipped entirely, not just the invalid settings.

 ❯ 1. Exit and fix manually
   2. Continue without these settings

 Enter to confirm · Esc to cancel

What Should Happen?

Error messages should be lucid and readable by a little old human

Error Messages/Logs

See above

Steps to Reproduce

I have no clue but it should not behave this way

Claude Model

Not sure / Multiple models

Is this a regression?

Yes, this worked in a previous version

Last Working Version

1.0.93

Claude Code Version

2.1.39

Platform

Anthropic API

Operating System

Ubuntu/Debian Linux

Terminal/Shell

iTerm2

Additional Information

_No response_