prodigy 0.4.4

Turn ad-hoc Claude sessions into reproducible development pipelines with parallel AI agents
Documentation 
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
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134

name: documentation-drift-parallel
mode: mapreduce

# Setup phase: Prepare documentation areas for parallel analysis
setup:
  # Create JSON file with documentation areas to analyze
  - shell: |
      cat > .prodigy/doc-areas.json << 'EOF'
      {
        "areas": [
          {"name": "README", "pattern": "README.md,**/README.md", "priority": "high"},
          {"name": "CLAUDE", "pattern": "CLAUDE.md,.claude/**/*.md", "priority": "high"},
          {"name": "commands", "pattern": ".claude/commands/*.md", "priority": "medium"},
          {"name": "workflows", "pattern": "workflows/*.yml,workflows/*.md", "priority": "medium"},
          {"name": "api_docs", "pattern": "src/**/*.rs", "priority": "high"},
          {"name": "examples", "pattern": "examples/**/*", "priority": "low"},
          {"name": "config", "pattern": "*.toml,*.yaml,*.yml", "priority": "medium"},
          {"name": "tests", "pattern": "tests/**/*.rs", "priority": "low"}
        ]
      }
      EOF
      echo "Created documentation areas configuration"
    capture_output: true

  # Initial full analysis to get baseline
  - claude: "/prodigy-docs-drift-analyze"
    commit_required: false
  
  - shell: |
      if [ -f .prodigy/documentation-drift.json ]; then
        cp .prodigy/documentation-drift.json .prodigy/drift-baseline.json
        echo "Baseline drift analysis saved"
      fi

# Map phase: Process each documentation area in parallel
map:
  # Input configuration - doc-areas.json contains areas array
  input: .prodigy/doc-areas.json
  json_path: "$.areas[*]"
  
  # Commands to execute for each documentation area (simplified syntax)
  agent_template:
    # Analyze drift for this specific area
    - claude: "/prodigy-docs-drift-analyze --area '${item.name}' --pattern '${item.pattern}'"
      commit_required: false
      timeout: 300

    # Fix critical issues in this area
    - claude: "/prodigy-docs-drift-update critical --area '${item.name}'"
      commit_required: true
      timeout: 300

    # Fix major issues in this area
    - claude: "/prodigy-docs-drift-update major --area '${item.name}'"
      commit_required: true
      timeout: 300

    # Capture metrics for this area
    - shell: |
        if [ -f .prodigy/documentation-drift.json ]; then
          jq --arg area "${item.name}" '{
            area: $area,
            priority: "${item.priority}",
            remaining_issues: .summary.total_issues,
            fixed_critical: .summary.fixed_critical // 0,
            fixed_major: .summary.fixed_major // 0
          }' .prodigy/documentation-drift.json > .prodigy/drift-${item.name}-result.json
        else
          echo '{"area": "${item.name}", "remaining_issues": 0}' > .prodigy/drift-${item.name}-result.json
        fi
      capture_output: true

  # Parallelization settings
  max_parallel: 4  # Run up to 4 agents in parallel

# Reduce phase: Aggregate results and finalize (simplified syntax)
reduce:
  # Fix any remaining minor issues globally
  - claude: "/prodigy-docs-drift-update minor"
    commit_required: true
    timeout: 300

  # Final analysis to measure improvements
  - claude: "/prodigy-docs-drift-analyze"
    commit_required: false

  # Generate comparison report
  - shell: |
      echo "=== Documentation Drift Resolution Summary ==="
      echo "Parallel agents processed: ${map.total}"
      echo "Successful: ${map.successful}"
      echo "Failed: ${map.failed}"
      echo ""

      if [ -f .prodigy/drift-baseline.json ] && [ -f .prodigy/documentation-drift.json ]; then
        baseline_total=$(jq '.summary.total_issues // 0' .prodigy/drift-baseline.json)
        final_total=$(jq '.summary.total_issues // 0' .prodigy/documentation-drift.json)
        fixed=$((baseline_total - final_total))

        echo "Baseline issues: $baseline_total"
        echo "Remaining issues: $final_total"
        echo "Issues fixed: $fixed"

        if [ "$final_total" -eq 0 ]; then
          echo "✅ All documentation drift has been resolved!"
        elif [ "$fixed" -gt 0 ]; then
          echo "✅ Fixed $fixed documentation issues"
        else
          echo "⚠️ Some documentation issues could not be fixed automatically"
        fi
      fi
    capture_output: true

  # Final lint and format
  - claude: "/prodigy-lint"
    commit_required: true

  # Test documentation examples
  - shell: "cargo test --doc 2>/dev/null || echo 'Note: Some doctests may need manual review'"

# This MapReduce workflow:
# 1. Sets up documentation areas for parallel processing
# 2. Analyzes and fixes drift in each area independently (in parallel)
# 3. Aggregates results and handles remaining issues
# 4. Provides comprehensive reporting on improvements
#
# Usage:
# prodigy cook workflows/documentation-drift-mapreduce.yml
#
# Benefits:
# - Parallel processing for faster execution
# - Area-based isolation reduces merge conflicts
# - Automatic retry on failures
# - Detailed progress tracking