LLM_code_benchmark/.github/workflows/codeql.yml at b98e9a0cee635a03a4a6532407fc7084bb62fd67 · TechNavii/LLM_code_benchmark · 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
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
name: CodeQL Analysis

on:
  push:
    branches: [main, master]
    paths:
      # Trigger on changes to source code only
      - 'server/**'
      - 'harness/**'
      - 'gui/**'
      - 'conftest.py'
  pull_request:
    branches: [main, master]
    paths:
      - 'server/**'
      - 'harness/**'
      - 'gui/**'
      - 'conftest.py'
  schedule:
    # Run weekly on Sundays at 3:00 AM UTC (after nightly scans)
    - cron: '0 3 * * 0'
  workflow_dispatch:
    # Allow manual trigger for debugging
  # NOTE: Do NOT use pull_request_target here - it runs with elevated permissions
  # from the base branch, which can be exploited by malicious fork PRs

# Only one CodeQL run per branch at a time
concurrency:
  group: codeql-${{ github.ref }}
  cancel-in-progress: true

# Permissions required for CodeQL and security tab
# Fork PRs: security-events: write is required for SARIF upload but GitHub
# restricts this for fork PRs - CodeQL handles this gracefully
permissions:
  security-events: write  # Required: upload SARIF results
  contents: read          # Required: checkout code
  actions: read           # Required: detect workflow changes

jobs:
  analyze:
    name: Analyze (${{ matrix.language }})
    runs-on: ubuntu-latest

    strategy:
      fail-fast: false
      matrix:
        # CodeQL supports: cpp, csharp, go, java, javascript, python, ruby, swift
        language: ['python', 'javascript']

    steps:
      - name: Checkout repository
        uses: actions/checkout@34e114876b0b11c390a56381ad16ebd13914f8d5  # v4

      # Initialize CodeQL tools for scanning
      - name: Initialize CodeQL
        uses: github/codeql-action/init@45c373516f557556c15d420e3f5e0aa3d64366bc  # v3
        with:
          languages: ${{ matrix.language }}
          # Configure paths to scan (consistent with meta.lintScope)
          # and paths to exclude (consistent with meta.excludePatterns)
          config: |
            paths:
              - server
              - harness
              - gui
            paths-ignore:
              - runs
              - .venv
              - .pytest_cache
              - '**/node_modules'
              - '**/__pycache__'
              - 'tasks/**/workspace'
              - '**/*.pyc'
          # Use security-and-quality queries for comprehensive analysis
          # Brownfield-safe: start with default queries, expand later
          queries: +security-and-quality

      # For Python, we don't need to build - CodeQL can analyze source directly
      # For JavaScript, CodeQL also analyzes source directly (no build needed)
      - name: Autobuild
        uses: github/codeql-action/autobuild@45c373516f557556c15d420e3f5e0aa3d64366bc  # v3

      # Perform CodeQL Analysis
      - name: Perform CodeQL Analysis
        uses: github/codeql-action/analyze@45c373516f557556c15d420e3f5e0aa3d64366bc  # v3
        with:
          category: "/language:${{ matrix.language }}"
          # Upload SARIF results to GitHub Security tab
          upload: true
          # Brownfield-safe: warn-only initially
          # To fail on high/critical alerts, add:
          # fail-on: error
          # For now, we report all findings without failing the job