builder.yaml

id: ai_project_builder_gemini
namespace: company.team
description: Generate a simple C++ project folder structure with starter files and README as a zip using Google Gemini (No Makefile)

inputs:
  - id: assignment_text
    type: STRING
    description: "The assignment description from the user."
    required: false

tasks:
  # 1️⃣ Generate project structure from assignment text using Gemini
  - id: generate_project_structure
    type: io.kestra.plugin.ai.completion.JSONStructuredExtraction
    provider:
      type: io.kestra.plugin.ai.provider.GoogleGemini
      apiKey: "{{ secret('GEMINI_API_KEY') }}"
      modelName: gemini-2.5-flash
    prompt: |
      # TASK
      You are a C++ project structure generator for university programming assignments. Generate a complete, compilable C++17 project with starter code and documentation.
      
      # OUTPUT FORMAT
      Return ONLY valid JSON. No markdown, no explanations, no code blocks - just raw JSON.
      
      {
        "project_name": "CamelCaseProjectName",
        "folders": ["src", "include", "tests"],
        "files": [
          {
            "path": "src/main.cpp",
            "starter_code": "#include <iostream>\n\nint main() {\n    // TODO: Implement core functionality\n    std::cout << \"Program started\" << std::endl;\n    return 0;\n}"
          },
          {
            "path": "include/MyClass.h",
            "starter_code": "#ifndef MYCLASS_H\n#define MYCLASS_H\n\nclass MyClass {\npublic:\n    MyClass();\n    void doSomething(); // TODO: Implement\n};\n\n#endif"
          }
        ],
        "readme": "# Project Title\n\n## Overview\nBrief description of the assignment.\n\n## Project Structure\n```\nproject_name/\n├── src/\n│   └── main.cpp\n├── include/\n│   └── MyClass.h\n└── README.md\n```\n\n## Files Description\n\n### src/main.cpp\nEntry point. Contains main() function and starter code.\n\n### include/MyClass.h\nHeader for a class used in the project.\n\n## Key Classes/Functions\n\n### MyClass\n- `MyClass()`: Constructor\n- `doSomething()`: TODO implementation\n\n## Compilation Instructions\n```bash\ng++ -std=c++17 -Wall -Iinclude src/*.cpp -o project_name\n./project_name\n```\n\n## Implementation Tasks\n1. Implement core functionality in main.cpp\n2. Implement MyClass methods\n3. Add unit tests if needed\n\n## Notes\n- Ensure code compiles without errors\n- Use only C++17 features\n- Keep starter code simple and documented"
      
      # CRITICAL RULES
      - "folders" MUST be a JSON array: ["src", "include"], not a string
      - "files" MUST be a JSON array of objects with "path" and "starter_code"
      - Use actual \n for line breaks in code
      - main.cpp is always required
      - Include optional headers in include/
      
      ASSIGNMENT:
      {{ trigger.body.assignment_text ?? inputs.assignment_text ?? "Create a simple C++ hello world program" }}
    jsonFields:
      - project_name
      - folders
      - files
      - readme
    schemaName: ProjectStructure

  # 2️⃣ Debug AI output
  - id: debug_ai_output
    type: io.kestra.plugin.core.log.Log
    message: |
      === GEMINI OUTPUT ===
      {{ outputs.generate_project_structure.extractedJson }}

  # 3️⃣ Fix and validate AI output
  - id: fix_and_validate_structure
    type: io.kestra.plugin.scripts.python.Script
    inputFiles:
      project_data.json: "{{ outputs.generate_project_structure.extractedJson }}"
    outputFiles:
      - "fixed_project_data.json"
    script: |
      import json, sys, os, re

      try:
          with open("project_data.json", "r") as f:
              data = json.load(f)

          # Fix folders if string
          if isinstance(data.get("folders", []), str):
              data["folders"] = json.loads(data["folders"])

          # Fix files if string
          if isinstance(data.get("files", []), str):
              try:
                  data["files"] = json.loads(data["files"])
              except json.JSONDecodeError:
                  # Replace unescaped quotes in starter_code
                  fixed_files = re.sub(r'(std::cout\s*<<\s*)"([^"]*)"', r"\1'\2'", data["files"])
                  data["files"] = json.loads(fixed_files)

          # Validation
          errors = []
          if not data.get("project_name"): errors.append("Missing project_name")
          if not isinstance(data["folders"], list) or not data["folders"]: errors.append("Invalid folders")
          if not isinstance(data["files"], list) or not data["files"]: errors.append("Invalid files")
          file_paths = [f.get("path") for f in data["files"] if isinstance(f, dict)]
          if "src/main.cpp" not in file_paths: errors.append("Missing src/main.cpp")
          if not data.get("readme"): errors.append("Missing readme")
          if errors:
              for e in errors: print(f"❌ {e}")
              sys.exit(1)

          with open("fixed_project_data.json", "w") as f:
              json.dump(data, f, indent=2)

      except Exception as e:
          print(f"❌ FATAL ERROR: {e}")
          sys.exit(1)

  # 4️⃣ Create files and zip
  - id: create_files_and_zip
    type: io.kestra.plugin.scripts.python.Script
    inputFiles:
      project_data.json: "{{ outputs.fix_and_validate_structure.outputFiles['fixed_project_data.json'] }}"
    outputFiles:
      - "project.zip"
    script: |
      import os, sys, json, zipfile

      try:
          with open("project_data.json", "r") as f:
              raw = f.read().strip()

              try:
                  data = json.loads(raw)
              except json.JSONDecodeError:
                  # Try to recover from invalid JSON by fixing common issues
                  import re

                  # Replace unescaped newlines inside strings
                  raw = raw.replace("\n", "\\n")

                  # Replace double quotes in C++ cout with single quotes
                  raw = re.sub(
                      r'(std::cout\s*<<\s*)"([^"]*)"',
                      r"\1'\2'",
                      raw
                  )

                  data = json.loads(raw)


          project_name = data["project_name"]
          project_path = os.path.join(os.getcwd(), project_name)
          os.makedirs(project_path, exist_ok=True)

          # Create folders
          for folder in data["folders"]:
              os.makedirs(os.path.join(project_path, folder), exist_ok=True)

          # Create files
          for file_info in data["files"]:
              full_path = os.path.join(project_path, file_info["path"])
              os.makedirs(os.path.dirname(full_path), exist_ok=True)
              with open(full_path, "w") as f:
                  f.write(file_info["starter_code"])

          # Create README.md
          with open(os.path.join(project_path, "README.md"), "w") as f:
              f.write(data["readme"])

          # Create zip
          zip_path = os.path.join(os.getcwd(), "project.zip")
          with zipfile.ZipFile(zip_path, "w", zipfile.ZIP_DEFLATED) as zipf:
              for root, dirs, files_in_dir in os.walk(project_path):
                  for file in files_in_dir:
                      zipf.write(os.path.join(root, file), os.path.relpath(os.path.join(root, file), project_path))

      except Exception as e:
          print(f"❌ FATAL ERROR: {e}")
          sys.exit(1)

outputs:
  - id: download
    type: FILE
    value: "{{ outputs.create_files_and_zip.outputFiles['project.zip'] }}"

  - id: project_info
    type: JSON
    value: "{{ read(outputs.fix_and_validate_structure.outputFiles['fixed_project_data.json']) }}"

triggers:
  - id: webhook
    type: io.kestra.plugin.core.trigger.Webhook
    key: umer