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
# Test execution with automatic debugging on failure
# Runs tests and automatically attempts to fix failures before proceeding to linting
- shell: "just test"
  on_failure:
    claude: "/prodigy-debug-test-failure --output ${shell.output}"
    max_attempts: 3
    fail_workflow: true  # Fail if tests can't be fixed after 3 attempts
    commit_required: true  # Create commits when fixing tests

- claude: "/prodigy-lint"  # Clean up code after tests pass