Add initial Copilot instructions and VSCode extension recommendations

.github/copilot-instructions.md

@@ -0,0 +1,56 @@
+# Rasa Codespaces Quickstart - AI Coding Agent Instructions
+
+## Project Overview
+This is a minimal Rasa Pro quickstart repository designed for GitHub Codespaces, providing a pre-configured development environment for conversational AI projects. The repository focuses on getting users up and running quickly with Rasa Pro in a containerized environment.
+
+## Environment Setup Critical Path
+1. **License Configuration**: Always check `.env` file first - the `RASA_LICENSE` environment variable is required for Rasa Pro functionality
+2. **Virtual Environment**: The devcontainer auto-creates `.venv/` via `uv venv` - always activate with `source .venv/bin/activate`
+3. **Environment Loading**: Run `source .env` to load license key before any Rasa commands
+4. **Package Installation**: Uses `uv` package manager, not standard pip - Rasa Pro installed from private index at `https://europe-west3-python.pkg.dev/rasa-releases/rasa-pro-python/simple`
+
+## Key Development Workflow
+```bash
+# Standard initialization sequence
+source .env                    # Load license
+source .venv/bin/activate     # Activate Python environment
+rasa init --template tutorial # Initialize project
+rasa train                    # Train model
+rasa inspect                  # Launch web interface
+```
+
+## Container Architecture
+- **Base Image**: `mcr.microsoft.com/devcontainers/python:1-3.10-bullseye`
+- **Package Manager**: `uv` for fast dependency resolution
+- **Python Version**: 3.10 (locked for Rasa compatibility)
+- **Port Forwarding**: Automatic via Codespaces for `rasa inspect` web interface
+- **Root User**: Container runs as root for simplified permissions
+
+## Project-Specific Patterns
+- **No Source Code**: This is a quickstart template - actual Rasa projects are created via `rasa init`
+- **Ephemeral Nature**: Designed for temporary exploration, not persistent development
+- **License-Dependent**: All Rasa commands fail without valid `RASA_LICENSE` in environment
+- **Tutorial Focus**: Default template is the Rasa tutorial bot for learning
+
+## Critical Files
+- `.env`: License key storage (must be manually populated)
+- `.devcontainer/devcontainer.json`: Complete environment definition
+- `README.md`: Step-by-step user instructions (authoritative workflow)
+
+## Custom Actions Setup (Rasa 3.10+)
+When working with generated Rasa projects, verify `endpoints.yml` contains:
+```yaml
+action_endpoint:
+  actions_module: "actions"
+```
+
+## VSCode Integration
+- **Recommended Extension**: Google Cloud Code for Rasa Pro integration
+- **Python Interpreter**: Auto-configured to `.venv/bin/python`
+- **Terminal Activation**: Python environment auto-activates in new terminals
+
+## Troubleshooting Common Issues
+- **License Errors**: Verify `RASA_LICENSE` is set and environment is sourced
+- **Package Not Found**: Ensure using `uv pip install` with Rasa Pro index URL
+- **Port Access**: Use notification popups in Codespaces for `rasa inspect` access
+- **Environment Issues**: Recreate with `uv venv && source .venv/bin/activate`

.vscode/extensions.json

@@ -0,0 +1,5 @@
+{
+    "recommendations": [
+        "googlecloudtools.cloudcode"
+    ]
+}