just-fucking-cancel

just-fucking-cancel

Analyze transactions, categorize subscriptions, generate HTML audit, help cancel.

Workflow

1. Get Transaction CSV

Ask user for bank/card CSV export. Common sources:

• Apple Card: Wallet → Card Balance → Export
• Chase: Accounts → Download activity → CSV
• Mint: Transactions → Export

2. Analyze Recurring Charges

Read CSV, identify recurring patterns:

• Same merchant, similar amounts, monthly/annual frequency
• Flag subscription-like charges (streaming, SaaS, memberships)
• Note charge frequency and total annual cost

3. Categorize with User

For each subscription, ask user to categorize:

• Cancel - Stop immediately
• Investigate - Needs decision (unsure, trapped in contract)
• Keep - Intentional, continue paying

Ask in batches of 5-10 to avoid overwhelming.

4. Generate HTML Audit

Copy template.html and populate:

• Update header summary:
  • Scope line: "found N subscriptions · N transactions"
  • Breakdown: "Cancelled N · Keeping N"
  • Savings: yearly amount big, monthly in parentheses
  • Timestamp: current date
• Add rows to appropriate sections (cancelled/investigate/keep)
• Include notes from user responses

Row templates in the HTML comments show the structure.

5. Cancel Subscriptions

When user checks items and copies from floating button, they'll paste: Cancel these: Service1 ($XX), Service2 ($XX)...

For each service:

1. Check common-services.md for cancel URL
2. Use browser automation to navigate and cancel
3. Update HTML row to cancelled status with date

HTML Structure

Three sections, auto-hide when empty:

• Cancelled (green badge, strikethrough) - Done items, the win
• Needs Decision (orange badge) - Has checkboxes for cancel selection
• Keeping (grey badge) - No checkboxes, just reference

Features:

• Floating copy button appears when items checked
• Privacy toggle blurs service names
• Collapsible sections via header click
• Dark mode support

Cancellation Tips

For difficult services, see common-services.md:

• Direct cancel URLs for 50+ services
• Dark pattern warnings (gym contracts, phone-only)
• Retention script responses
• Credit card dispute backup

Privacy

All data stays local. Transaction CSVs are analyzed in-session only.