Majk ·
Step 1 of 12
💬
The conversation will appear here as you step through →
Y
You

Read all of these credit card statements and build me a financial dashboard as a single self-contained HTML file. I want: a spending breakdown by category with a visual chart, a month-by-month trend showing how my total spend changes over time, a subscription audit table, my top 10 biggest individual transactions, and a summary section with the key things I should actually pay attention to. Make it look good — something I'd be comfortable showing someone.

11:04 AM
A
AI Assistant Majk Rabbit

I'll read all 18 statements first, then categorise and synthesise the data, then build the dashboard.

📄 Read statement_jan_2023.pdf — extracting transactions, amounts, merchants
📄 Read statement_feb_2023.pdf — extracting transactions, amounts, merchants
📄 Read statement_mar_2023.pdf — extracting transactions, amounts, merchants
📄 Read … 15 more statements
>_ Bash python3 categorise.py — grouping merchants into categories, calculating monthly totals
>_ Bash python3 find_recurring.py — identifying subscriptions and recurring charges
✎ Write financial_dashboard.html — building dashboard with inline CSS and SVG charts
11:04 AM
financial_dashboard.html
💰 Financial Dashboard
Jan 2023 – Jun 2024 · 18 months · 2,340 transactions
$47,820
Total Spend
$2,657
Monthly Avg
$1,661
Subscriptions / yr
$2,840
Largest Transaction
Spending by Category
Dining
28%
Subscriptions
18%
Travel
16%
Groceries
14%
Shopping
13%
Other
11%
Monthly Spend — Jan 2023 to Jun 2024
J
F
M
A
M
J
J
A
S
O
N
D
J
F
M
A
M
J
Active Subscriptions
ServiceAmountAnnualNote
Adobe CC$54.99/mo$659/yrlargest
New York Times$17.00/mo$204/yr↑ 70%
Netflix$15.49/mo$185/yr↑ price
Dropbox Plus$11.99/mo$143/yroverlap
Spotify$10.99/mo$131/yrkeep
⚠️ 3 Things to Act On
  • Two cloud storage services (iCloud + Dropbox) — $178/yr combined, probably redundant
  • New York Times raised its price 70% since you signed up — now $17/mo
  • Dining trending 22% above your 2023 average over the last 4 months
Y
You

I have a folder with my Q3 project notes, meeting docs, retros, and a rough bullet list of things I shipped. Build me a polished one-page HTML summary of the quarter — headline achievements at the top, impact numbers where you can find them, a timeline of what shipped, and a section on what I'm carrying into Q4. Make it look like something I'd actually send to my manager.

3:41 PM
A
AI Assistant Majk Rabbit

I'll read through all the source files to find the achievements, numbers, and timeline, then build the page.

📄 Read q3_projects.md — scanning for shipped items, outcomes, impact metrics
📄 Read meeting_notes_jul_aug_sep.md — extracting decisions, milestones, blockers resolved
📄 Read retro_q3.md — extracting what went well, what shipped early or late
📄 Read goals_q3.md — cross-referencing shipped work against stated goals
✎ Write q3_summary.html — building one-pager with achievements, timeline, Q4 setup
3:41 PM
q3_summary.html
📊 Q3 2024 — Work Summary
Sarah Chen · Engineering · July – September 2024
Headline Achievements
🚀
API redesign shipped 2 weeks ahead of schedule despite scope increase mid-quarter
📚
Internal docs adopted by 3 other teams — now the org-wide reference for the API
⚡
New-hire onboarding time reduced 40% following API redesign and new documentation
Impact
40%
Faster onboarding
3
Teams adopted docs
0
P0 incidents post-launch
Timeline
Jul
API Design
Aug
Build + Review
Sep
Launch + Adoption
Into Q4
→
Cross-team coordination work — started in Q3, not yet fully realised
→
Technical debt item deferred from Q3 — needs a plan
Y
You

Here are 18 months of credit card statements. Build me a dashboard that gives me a genuinely useful picture of my finances — you decide what sections and charts would actually be valuable. Single HTML file.

2:15 PM
A
AI Assistant Majk Rabbit

I'll read through everything first, then decide what structure would actually be useful before building.

📄 Read statement_jan_2023.pdf — reading all transactions
📄 Read statement_feb_2023.pdf — reading all transactions
📄 Read … 16 more statements
>_ Bash python3 analyse.py — finding categories, trends, anomalies, and what changed over time
✎ Write financial_dashboard.html — building around what the data actually shows
2:15 PM
A
AI Assistant Majk Rabbit

Here's how the two approaches compare on the same data:

Method promptingOutcome prompting
StructureYour 5 sections, exactly as specified6 sections — added one you didn't ask for
Added section—"Spending drift" — shows how each category has shifted vs your personal baseline, not just raw totals
Summary callout3 items flagged2 items flagged — dining and cloud storage — but with a clearer explanation of why they matter
What you getExactly what you specified, predictablySomething designed around what the data is actually telling you

The "spending drift" section wouldn't have appeared in the method version because you didn't ask for it — but it's probably the most useful thing in the dashboard.

2:18 PM · 4,450 tokens
Y
You

Can you add a note at the top explaining what this is, so I can send it to someone who doesn't have context?

11:09 AM
A
AI Assistant Majk Rabbit

Added a header section to financial_dashboard.html with a one-paragraph explanation: what the data covers (18 months of personal credit card statements, Jan 2023 – Jun 2024), how it was generated, and a note that all figures are approximate due to merchant categorisation. Ready to share.

11:09 AM · 312 tokens
Y
You

The dining spend looks high — can you break that down by restaurant and show me which months were the outliers?

11:10 AM
A
AI Assistant Majk Rabbit

Dining breakdown across 18 months:

  • Top merchants: Uber Eats ($1,240 total), local restaurants cluster ($890), one-off business dinners ($620)
  • Outlier months: October 2023 (+$380 above your average — two business dinners and a run of Uber Eats in the same week) and March 2024 (+$290 — travel month, eating out every day)
  • Baseline: Your "normal" dining month is around $410. Everything else is noise around that.

Added a dining drilldown section to the dashboard with a per-merchant bar chart and the outlier months highlighted.

11:10 AM · 680 tokens