Agent skill
scorecard-update
Review Azure Data Explorer dashboards for Copilot CLI metrics and update Google Sheets scorecard. Connects to existing Chrome session for authentication.
Install this agent skill to your Project
npx add-skill https://github.com/patniko/skills/tree/main/scorecard-update
Metadata
Additional technical details for this skill
- author
- patrick
- version
- 1.0
SKILL.md
Scorecard Update
Review Copilot CLI metrics from Azure Data Explorer dashboards and update the Google Sheets scorecard.
Prerequisites
Before using this skill, start Chrome with remote debugging:
# macOS
/Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome --remote-debugging-port=9222
# Then log into your Azure and Google accounts in that browser session
Workflow
This skill follows a review-before-update workflow:
- Read scorecard → Identify which Copilot CLI metrics need updating
- Capture dashboards → Screenshot and extract data from Azure Data Explorer
- Generate summary → Create a review document with proposed updates
- Human review → You approve/modify the proposed changes
- Update scorecard → Apply approved changes to Google Sheets
Data Sources
Dashboards (Azure Data Explorer)
| Dashboard | Purpose |
|---|---|
| Launch Metrics | Metrics since launch date |
| 30-Day Trends | Rolling 30-day window |
| Overview | General dashboard view |
Scorecard (Google Sheets)
Quick Start
# 1. Ensure Chrome is running with debugging port (see Prerequisites)
# 2. Install dependencies (first time only)
cd scorecard-update && npm install
# 3. Run the workflow
node scripts/workflow.js
Step-by-Step Instructions
Step 1: Read the Scorecard
Connect to Chrome and navigate to the scorecard to identify metrics that need updating:
// The script will:
// 1. Open the scorecard
// 2. Find the Copilot CLI section
// 3. List metrics and their current values
// 4. Identify which need updating (blank, outdated, etc.)
Step 2: Capture Dashboard Data
For each dashboard, the script will:
- Navigate to the URL
- Wait for charts to render
- Take a screenshot
- Extract visible metrics/numbers from the page
Step 3: Review Summary
Before any updates, you'll receive a summary like:
## Proposed Scorecard Updates
| Metric | Current Value | New Value | Source |
|--------|---------------|-----------|--------|
| DAU | 1,234 | 1,456 | 30-Day Dashboard |
| ... | ... | ... | ... |
Screenshots saved to: ./screenshots/
Please review and confirm before updating.
Step 4: Update Scorecard
After your approval, the script navigates to the scorecard and updates the specified cells.
Scripts
| Script | Purpose |
|---|---|
workflow.js |
Main orchestrator - runs full workflow |
connect.js |
Connect to Chrome debugging session |
capture-dashboards.js |
Screenshot and extract dashboard data |
read-scorecard.js |
Read current scorecard values |
update-scorecard.js |
Write updates to scorecard |
Output Files
scorecard-update/
├── screenshots/ # Dashboard screenshots (timestamped)
├── data/
│ ├── scorecard-current.json # Current scorecard state
│ ├── dashboard-data.json # Extracted dashboard metrics
│ └── proposed-updates.json # Summary for review
└── summaries/
└── YYYY-MM-DD-summary.md # Human-readable summary
Notes
- All changes require human approval before applying
- Screenshots provide audit trail of data sources
- The script waits for Azure dashboards to fully render (they can be slow)
- If a metric can't be automatically extracted, it will be flagged for manual review
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
issue-labeler
Analyze unlabeled GitHub issues and generate label recommendations for review. Supports batch submission after approval.
canvas
Spawn interactive terminal TUI components (calendars, documents, flight bookings) with real-time IPC communication. Display rich content and collect user selections in tmux split panes.
shipment-tracker
Generate a visual report of what was shipped in GitHub repositories over a specified time period by analyzing merged PRs and closed issues.
meeting-to-issues
Parse meeting transcripts to extract action items and create GitHub issues after user confirmation. Converts discussions into trackable work items.
issue-triage
Fetch GitHub issues and use LLM judgment to prioritize them based on importance, clarity, delegation potential, and urgency. Helps identify what to work on next.
competitive-analysis
Compare local codebase capabilities against competitor products by researching competitor features via web search and documentation, analyzing local code patterns, and generating an interactive HTML comparison report.
Didn't find tool you were looking for?