cora-code 0.18.0

CLI-first AI code review — BYOK, diff/scan/branch, pre-commit hooks
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51

name: clean-code
description: "Broad quality — readability, naming, complexity — best for team projects"
version: "1.0"

focus_areas:
  - id: readability
    weight: 8
    action: warn
    rules:
      - "Functions should do one thing — flag long functions (>50 lines)"
      - "Use descriptive names — avoid single-letter vars (except loop counters)"
      - "Avoid magic numbers — extract to named constants"
      - "Keep nesting depth reasonable (max 3-4 levels)"

  - id: complexity
    weight: 7
    action: warn
    rules:
      - "Flag high cyclomatic complexity — too many branches in one function"
      - "Avoid deeply nested conditionals — use early returns or guard clauses"
      - "Break large modules into focused submodules"

  - id: error_handling
    weight: 8
    action: warn
    rules:
      - "Handle errors explicitly — no silent swallowing or empty catch blocks"
      - "Use appropriate error types — don't return generic errors everywhere"
      - "Propagate errors with context (wrap with 'failed to X')"

  - id: dry
    weight: 6
    action: warn
    rules:
      - "Flag duplicated code — suggest extraction into shared functions"
      - "Avoid copy-paste with minor variations — parameterize instead"

  - id: testing
    weight: 5
    action: info
    rules:
      - "Suggest tests for complex logic (edge cases, error paths)"
      - "Flag untested public APIs"

ignore_areas: []

review_style:
  tone: standard
  detail_level: standard
  suggest_fixes: true
  max_findings: 30