open-code-review/action.yml at main · raye-deng/open-code-review · GitHub

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
name: 'Open Code Review'
description: 'AI code quality gate — detect hallucinations, stale APIs, logic gaps in PRs'
author: 'raye-deng'

branding:
  icon: 'shield'
  color: 'purple'

inputs:
  sla:
    description: 'Scan level: L1 (structural), L2 (+ embedding), L3 (+ LLM deep scan)'
    required: false
    default: 'L1'
  threshold:
    description: 'Minimum score to pass (0-100)'
    required: false
    default: '70'
  scan-mode:
    description: 'Scan mode: diff (changed files only) or full'
    required: false
    default: 'diff'
  ollama-url:
    description: 'Ollama API URL for local AI (L2/L3). If not provided, L2 uses TF-IDF fallback.'
    required: false
  fail-on-low-score:
    description: 'Fail workflow if score is below threshold'
    required: false
    default: 'true'
  github-token:
    description: 'GitHub token for PR comments'
    required: false
    default: '${{ github.token }}'

outputs:
  score:
    description: 'Quality score (0-100)'
  grade:
    description: 'Letter grade (A-F)'
  issues-count:
    description: 'Total issues found'

runs:
  using: 'node20'
  main: 'packages/github-action/dist/index.js'