This starter repo does four things:

1. Lets you create GitHub issues that auto-assign @copilot
2. Gives Copilot repository-specific instructions
3. Prepares a lightweight environment for Copilot cloud agent
4. Deploys a static dashboard to GitHub Pages with freshly generated synthetic data on each deploy

Domain: customer support operations for a B2B SaaS company

Embedded hypotheses in the generated data:

1. Higher backlog today increases SLA breach risk tomorrow
2. Email tickets have slower first response and lower CSAT than chat
3. Weekend under-staffing creates a Monday ticket spike
4. APAC has the strongest Monday backlog effect because staffing is leaner

The dashboard reads site/data/dashboard.json and renders:

• KPI cards
• trend charts
• channel and region comparisons
• hypothesis callouts
• alert panels

1. Create a GitHub repo and copy these files in.
2. Enable GitHub Copilot cloud agent for the repo.
3. In Settings → Pages, choose GitHub Actions as the source.
4. Push to your default branch.
5. Open Actions and run Deploy GitHub Pages dashboard once if needed.
6. Create a new issue using the Copilot Bug Fix template.

If your plan and repo permissions allow it, the issue template auto-assigns @copilot, and Copilot can open a branch and PR for review.

python scripts/generate_data.py --seed 42 --days 180 --output site/data/dashboard.json
python -m http.server 8000

Then open http://localhost:8000/site/.

• The dashboard is intentionally static and dependency-free.
• The dataset is synthetic but structured to feel realistic.
• Each GitHub Pages deploy uses a fresh seed, so the site changes slightly from deploy to deploy.