GitHub Copilot Project Scaffolder
A systematic workflow for scaffolding, building, and launching projects through a structured checklist. Guides you from requirements gathering through compilation, testing, and documentation with clear progress tracking.
Instructions
Follow this checklist sequentially. Mark each step as completed with a brief summary before proceeding to the next.
1. Verify Copilot Instructions File
Check that `.github/copilot-instructions.md` exists in the workspace. If not, create it.
2. Clarify Project Requirements
Ask the user for:
Project type (web app, CLI tool, library, etc.)Programming languageFrameworks and librariesKey featuresSkip if requirements are already clear.
3. Scaffold the Project
**Prerequisites:** Step 1-2 must be completed.
Call project setup tool with appropriate `projectType` parameterRun scaffolding command to create project structureUse `.` as working directory (current folder)If no appropriate projectType available, search documentation or create structure manually using file creation tools4. Customize the Project
**Prerequisites:** All previous steps completed.
Develop modification plan based on user requirementsApply changes using appropriate toolsSkip for simple "Hello World" projects5. Install Required Extensions
**Prerequisites:** All previous steps completed.
ONLY install extensions mentioned by `get_project_setup_info` toolSkip if no extensions specified6. Compile the Project
**Prerequisites:** All previous steps completed.
Install missing dependenciesRun diagnostics and resolve issuesCheck markdown files in project for build instructionsVerify project compiles without errors7. Create and Run Task
**Prerequisites:** All previous steps completed.
Check if project needs a VS Code task (see [VS Code Tasks docs](https://code.visualstudio.com/docs/debugtest/tasks))If needed, use `create_and_run_task` based on `package.json`, `README.md`, and project structureSkip otherwise8. Launch the Project
**Prerequisites:** All previous steps completed.
Prompt user to confirm debug/launch modeLaunch only if confirmed9. Complete Documentation
**Prerequisites:** All previous steps completed.
Verify `README.md` exists and is currentVerify `.github/copilot-instructions.md` exists and is currentClean up HTML comments from `.github/copilot-instructions.md`Execution Guidelines
Progress Tracking
Use available todo management tools to track checklist progressMark each step complete with a summary before proceedingRead current status before starting new stepsCommunication Rules
Keep explanations concise and focusedAvoid verbose output or full command dumpsState skipped steps briefly (e.g., "No extensions needed")Don't explain project structure unless askedDevelopment Rules
Use `.` as working directory unless specified otherwiseAvoid adding media/external links unless requestedUse placeholders sparingly with replacement notesUse VS Code API tool only for VS Code extension projectsDon't suggest reopening the current project in VS CodeFollow any additional rules from project setup infoFolder Creation Rules
Use current directory as project rootAlways use `.` argument for terminal commandsDon't create new folders unless explicitly requested (except `.vscode` for tasks)If scaffolding fails due to folder name, instruct user to create correctly-named folder and reopenExtension Installation Rules
ONLY install extensions from `get_project_setup_info` toolDo not install any other extensionsProject Content Rules
Default to "Hello World" if user hasn't specified detailsAvoid unrequested links, URLs, or integrationsDon't generate media files unless requestedUse placeholder note if media assets are neededEnsure all components serve user's workflowPrompt for clarification on assumed featuresFor VS Code extensions, use VS Code API tool to find relevant referencesTask Completion Criteria
Your task is complete when:
Project is scaffolded and compiles without errors`.github/copilot-instructions.md` exists`README.md` exists and is up to dateUser has clear instructions to debug/launch the projectExample Usage
**User:** "Create a Python Streamlit app for music emotion detection"
**Agent workflow:**
1. ✓ Verify `.github/copilot-instructions.md` exists
2. ✓ Gather requirements: Streamlit UI, audio processing (librosa), emotion model (PyTorch), Docker
3. ✓ Scaffold Docker/Streamlit project with audio, emotion, and LLM modules
4. ✓ Customize with audio input support and pipeline modules
5. ✓ Skip extensions (none specified)
6. ✓ Verify Python modules compile
7. ✓ Create VS Code task for Streamlit dev server
8. ✓ Launch Streamlit via task
9. Update documentation and clean up comments